TW202321130A - Freight management system and method thereof - Google Patents

Abstract

Example freight management systems and methods are described. In one implementation, techniques receive at least one wide angle camera image from a sensor tower, where the sensor tower is located proximate a loading dock and the wide angle camera image is associated with at least a portion of the loading dock. The techniques also receive multiple high precision camera images from the sensor tower, where the plurality of high precision camera images are associated with at least a portion of the loading dock. The techniques process the wide angle image using a first convolutional neural network (CNN) and process the multiple high precision images using a second CNN. The techniques identify a freight item proximate the loading dock based on the processed high precision images.

Description

Cargo management system and method thereof

The present disclosure relates to managing warehouse operations, including shipping and receiving goods via docks.

Warehouses and other buildings that ship and receive items typically use loading docks to transfer those items between the warehouse and the trucks that ship and receive any number of items. In some cases, forklifts and other equipment can unload items from trucks into warehouses and can load items from warehouses onto trucks.

It is important to track the movement of these items, verify that received shipments contain the correct items, check received shipments for damaged items, check for evidence of tampered items, etc. Typically, these activities are performed manually by human operators who inspect incoming shipments and compare them to bills of lading or other tracking systems. These manual operations are often slow and error-prone.

The present invention discloses a cargo management system, comprising: one or more processors; and one or more non-transitory computer-readable media storing instructions executable by the one or more processors, wherein the instructions are executed when executed Causing the system includes receiving at least one wide-angle camera image from a sensor tower, wherein the sensor tower is located adjacent to a dock and the wide-angle camera image is associated with at least a portion of the dock; The sensor tower receives a plurality of high-precision camera images, wherein the plurality of high-precision camera images is associated with at least a portion of the dock; preprocessing the wide-angle camera image and the plurality of high-precision camera images to generate a pre-processed wide-angle image and a plurality of pre-processed high-resolution images; process the pre-processed wide-angle image using a first convolutional neural network (CNN); and process the complex number using a second convolutional neural network (CNN) A preprocessed high-resolution image.

The cargo management system further includes identifying at least one device operating near the loading dock based on processing the plurality of pre-processed high-resolution images using a second CNN.

The system performs operations further comprising identifying a cargo item approaching the dock based on processing the plurality of pre-processed high-resolution images using a second CNN. The system performs operations further comprising identifying a plurality of objects associated with the cargo item based on processing the plurality of pre-processed high-resolution images using the second CNN. The operations further include receiving manual input based on a result of processing the plurality of pre-processed high-resolution images using the second CNN. The operations also include training the second CNN based on the received artificial input. The operations further comprise analyzing the pre-processed high-resolution imagery to identify whether at least one item associated with the cargo item is damaged or to identify whether at least one event associated with the cargo item has been tampered with. The operation also includes determining whether the cargo item is being loaded or unloaded by the loading dock. The operations also include identifying at least one piece of equipment operating proximate to the dock. The operations further include detecting an informative texture associated with at least one item proximate to the dock based on processing the plurality of pre-processed high-resolution images using the second CNN. The operations further include transmitting the information texture detected to be associated with at least one object proximate to the dock to a remote computing system. The operation further includes associating the informative texture with a particular product. The operation further includes integrating the specific product data with higher level logistics data.

The cargo management system described above, wherein identifying the cargo item includes analyzing the label on the cargo item, analyzing the text on the cargo item, analyzing the logo on the cargo item, analyzing the size of the cargo item, and analyzing the color of the cargo item Or analyze at least one of the volumes of the item of cargo.

The present invention further discloses a cargo management method, comprising the following steps: receiving at least one wide-angle camera image from a sensor tower, wherein the sensor tower is located near a loading and unloading dock and the at least one wide-angle camera image is consistent with the loading and unloading dock. at least a portion is associated; receiving a plurality of high precision camera images from the sensor tower, wherein the plurality of high precision camera images are associated with at least a portion of the dock; using a first convolutional neural network (CNN ) processing the wide-angle image; processing the plurality of high-resolution images using a second convolutional neural network (CNN); and identifying cargo items near the loading dock according to the processed plurality of high-resolution images.

The cargo management method further includes preprocessing the at least one wide-angle camera image and the plurality of high-precision camera images to generate a pre-processed wide-angle image and a plurality of pre-processed high-precision images. The cargo management method further includes identifying a plurality of objects associated with the cargo item based on processing the plurality of high-resolution images using a second CNN. wherein identifying the item of cargo includes analyzing a label on the item of cargo, analyzing text on the item of cargo, analyzing a logo on the item of cargo, analyzing the size of the item of cargo, analyzing the color of the item of cargo, or analyzing the volume of the item of cargo at least one of the It further includes analyzing the plurality of processed high-resolution images to identify whether at least one item associated with the cargo item is damaged or to identify whether at least one item associated with the cargo item has been tampered with.

The cargo management method further includes training a second CNN based on receiving manual input associated with the processed high-precision images.

This application claims priority to US Application Nos. 17/488,031 and 17/488,033, filed September 28, 2021, the entire contents of which are incorporated herein by reference for all purposes.

In some embodiments, the systems and methods discussed herein perform various activities associated with warehouse operations, such as incoming and outgoing goods and the like. In particular embodiments, these systems and methods are associated with identifying, analyzing, and tracking items of cargo that move through docks at a warehouse or other facility.

In the following disclosure, reference is made to the accompanying drawings which form a part hereof, and in which are shown by way of example specific embodiments in which the disclosure may be practiced. It is to be understood that other embodiments may be utilized and structural changes may be made without departing from the scope of the present disclosure. References in the specification to "one embodiment," "an embodiment," "an exemplary embodiment," etc. mean that the described embodiments may include a particular feature, structure, or characteristic, but each embodiment does not necessarily have to. including the particular feature, structure or characteristic described. Moreover, such expressions are not necessarily referring to the same embodiment. Furthermore, when a particular feature, structure or characteristic is described in conjunction with an embodiment, whether explicitly described or not, it is considered within the knowledge of those skilled in the art to affect such feature, structure or characteristic in combination with other embodiments.

Embodiments of the systems, devices, and methods disclosed herein may include or utilize a special purpose or general purpose computer, including computer hardware such as one or more processors and system memory, as discussed herein. Embodiments within the scope of the present disclosure may also include physical and other computer-readable media for carrying or storing computer-executable instructions and/or data structures. Such computer-readable media can be any available media that can be accessed by a general purpose or special purpose computer system. A computer-readable medium that stores computer-executable instructions is a computer storage medium (devices). Computer-readable media that carry computer-executable instructions are transmission media. Thus, by way of example and not limitation, embodiments of the present disclosure may include at least two distinct types of computer-readable media: computer storage media (devices) and transmission media.

Computer storage media (devices) include RAM, ROM, EEPROM, CD-ROM, solid state drives (“SSD”) (e.g., RAM-based), flash memory, phase change memory (“PCM”), other types of memory, other optical disk storage, disk storage, or other magnetic storage device, or any other medium that can be used to store the desired program code means in the form of computer-executable instructions or data structures, and can be accessed by a general-purpose or special-purpose computer.

The implementation of the devices, systems and methods disclosed in this paper can communicate through computer networks. A "network" is defined as one or more data links capable of transmitting electronic data between computer systems and/or modules and/or other electronic devices. When information is transmitted or provided to a computer over a network or other communication connection (wired, wireless, or a combination of wires and wireless), the computer correctly sees the connection as the transmission medium. Transmission media can include network and/or data links which can be used to carry desired program code means in the form of computer-executable instructions or data structures and which can be accessed by a general purpose or special purpose computer. Combinations of the above should also be included in the scope of computer-readable media.

Computer-executable instructions include, for example, instructions and materials that, when executed at a processor, cause a general-purpose computer, special-purpose computer, or special-purpose processing device to perform a particular function or functional group of functions. Computer-executable instructions may be, for example, binary files, instructions in an intermediate format such as assembly language, or even source code. Although subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that subject matter defined in the claims is not necessarily limited to the features or acts described herein. Rather, the features and acts described in the present disclosure are described as example forms of implementing the claims.

Those skilled in the art will appreciate that the present disclosure can be practiced in network computing environments having many types of computer system configurations, including personal computers, desktop computers, notebook computers, message processors, handheld devices, multiprocessor Systems, microprocessor-based or programmable consumer electronics, network PCs, minicomputers, mainframe computers, mobile phones, PDAs, tablet computers, pagers, routers, switches, various storage devices, etc. The present disclosure can also be practiced in a distributed system environment in which local and remote computer systems linked by a network (either by a fixed-wire data link, a wireless data link, or by a combination of fixed-wire and wireless data links) are both perform tasks. In a distributed system environment, program modules may be located in both local and remote memory storage devices.

Furthermore, where appropriate, the functions described herein may be performed in one or more of: hardware, software, firmware, digital components, or analog components. For example, one or more application specific integrated circuits (ASICs) may be programmed to implement one or more of the systems and procedures described herein. Certain terms are used throughout the description and claims to refer to particular system components. Components may be referred to by different names, as will be understood by those skilled in the art. This document does not intend to distinguish between components that have different names but have the same function.

It should be noted that sensor embodiments discussed herein may include computer hardware, software, firmware, or any combination thereof to perform at least a portion of their functions. For example, a sensor may include program code configured to execute in one or more processors, and may include hardware logic/circuitry controlled by the program code. These example devices are provided herein for purposes of illustration and are not intended to be limiting. Embodiments of the present disclosure may be implemented in other types of devices, as known to those skilled in the relevant art.

At least some embodiments of the present disclosure are directed to a computer program product including such logic (eg, in software) stored on any computer usable medium. When executed in one or more data processing devices, such software causes the devices to operate as described herein.

"FIG. 1" is a block diagram illustrating an environment 100 in which example embodiments may be implemented. As shown in FIG. 1 , the first sensor tower 104 and the second sensor tower 106 are located on opposite sides of the dock 102 . In some embodiments, environment 100 is a warehouse, manufacturing facility, sorting facility, or any other facility having at least one loading dock. Although two sensor towers 104, 106 are shown in "Fig. 1". As shown in FIG. 1 , other embodiments may include a single sensor tower located on one side of the dock 102 . In some embodiments, the loading dock 102 may have doors (not shown) to control access to the loading dock 102 from outside the facility.

Any number of cargo items 108 may pass through dock 102, such as cargo items 108 are loaded onto trucks or other vehicles, and cargo items 108 are received from trucks or other vehicles. Cargo items 108 may include any type of container, collection of items, or other items that may be shipped and passed through dock 102 . Cargo items 108 include palletized items, racks holding one or more items, single bulky items (such as large containers of liquid or dry food items, oil drums, and vehicle engines), shrink-wrapped or bundled together multiple items wait. In some embodiments, forklifts, robots, or other machines move cargo items 108 through dock 102 when unloading or loading trucks or other vehicles.

In operation, cargo item 108 is moved across dock 102 at the normal operating speed of a forklift or other machine moving the cargo item. Sensor towers 104 and 106 have multiple cameras that can instantly scan cargo items 108 as forklifts or other machines move across dock 102 at their normal operating speed. Accordingly, sensor towers 104 and 106 may be installed on existing docks 102 without interrupting normal operations for receiving and delivering cargo items 108 . As discussed in greater detail herein, sensor towers 104 and 106 include a plurality of cameras that passively scan cargo items 108 as they pass through dock 102 .

In some embodiments, specific sensor towers can scan various cargo items without requiring reconfiguration between different types of cargo items. For example, one or more sensor towers may scan a pallet packaged item, then a single container item, then a shelf item. These different types of cargo items can be continuously scanned by one or more sensor towers without any changes to the sensor towers. Thus, an adaptable sensor tower allows loading and unloading of cargo regardless of the various types of cargo associated with a particular truck.

As shown in FIG. 1 , sensor towers 104 and 106 may be coupled to other devices and systems via data communication network 110 . In some embodiments, the data communication network 110 includes any type of network topology using any communication protocol. In addition, the data communication network 110 may include a combination of two or more communication networks. In some embodiments, the data communication network 110 includes a cellular communication network, the Internet, a local area network, a wide area network, or any other communication network.

In the example of environment 100, sensor towers 104 and 106 are coupled to interface with one or more wearable devices 112, one or more robotic devices 114, one or more forklifts 116, one or more warehouse devices 118, One or more operating platforms 120 , and one or more cloud-based computing systems 122 . Wearable device 112 may include any device worn by a warehouse worker (eg, a forklift operator) to provide feedback (eg, auditory, visual, or tactile feedback) related to one or more warehouse activities. Example information that may be communicated to warehouse personnel via wearable device 112 includes time differences in items passing sensor towers 104 and 106, items being shipped to or from, identification of visual defects in one or more items, unreadable , or one of the sensor towers 104 and 106 fails to identify items associated with the cargo.

Robotic devices 114 may include any device that assists in facility operations, such as moving objects, scanning objects, locating cargo items, and the like. Forklifts 116 may include automated forklifts or manual forklifts. Warehouse equipment 118 may include any other equipment that manages logistics within a warehouse and any other tasks required to manage inventory, shipping, receiving, scheduling, and the like. Operations platform 120 may provide various logistical and operational information to one or more systems and personnel. Additional details regarding operating platform 120 are discussed herein. The cloud-based computing system 122 can include any number of computing devices, such as servers, that can perform various tasks, activities, data storage, and the like. As discussed herein, cloud-based computing system 122 may perform analysis of imagery captured by sensor towers 104 and 106 and identification of one or more objects associated with cargo item 108 .

As discussed herein, each sensor tower 104 and 106 includes an independent computing device that performs various tasks, such as analyzing captured imagery and identifying one or more objects associated with cargo item 108 . In some embodiments, sensor towers 104 and 106 may also access one or more cloud-based computing systems 122 to perform at least some analysis of captured images and identification of items associated with cargo item 108 . In other embodiments, sensor towers 104 and 106 may not include their own computing devices. In such cases, sensor towers 104 and 106 may rely on one or more cloud-based computing systems 122 to perform analysis of captured imagery and identification of objects associated with cargo item 108 .

In another embodiment, sensor tower 104 has its own computing device that analyzes the captured imagery and identifies one or more objects associated with cargo item 108 . However, in this embodiment, sensor tower 106 does not have its own computing device. In this case, computing devices in sensor tower 104 can analyze images captured by sensor tower 106 and identify objects in the images captured by sensor tower 106 . For example, sensor tower 106 may transmit its captured imagery to sensor tower 104 for processing.

In some embodiments, the systems and methods discussed herein can determine the presence or absence of an item in an item of cargo, and detect physical damage, leakage, and tampering with the item or item of cargo. The described systems and methods can also determine the size of a cargo item, detect the type of packaging information (eg, expiration date or logo), and sense the temperature of the object or cargo item. These systems and methods can further identify items or cargo items that do not have standard labels, barcodes, or other identifiers. Additionally, the described systems and methods can track time spent loading cargo onto trucks, unloading cargo from trucks, and other dock-related operations.

"FIG. 2" illustrates an embodiment of a sensor tower 200 including eleven cameras numbered 202-222. Sensor tower 200 may include at least one housing that provides support for cameras 202-222 and other components associated with sensor tower 200, as discussed herein. The housing can have any shape, any number of parts, and can be made of any one or more materials. Sensor towers 104 and 106 may be similar or identical to sensor tower 200 . In some embodiments, cameras 202-222 may be different types of cameras, such as wide-angle cameras and high-precision cameras. Cameras 202-222 may operate in the visible spectrum or any other spectrum. In some embodiments, the cameras 202-222 may have different lenses based on the size (eg, width) of the dock. In some embodiments, at least one of the cameras 202-222 may be a monochrome sensor and at least one of the other cameras 202-222 is a color camera. In other implementations, different cameras 202-222 may have different numbers of pixels or different pixel sizes. In particular embodiments, one or more of the cameras 202-222 may be an IR (infrared) camera, a 3D (three-dimensional) time-of-flight camera, a thermal imaging camera, or the like.

In some embodiments, the size of the dock 102 and the distance between the sensor towers and the distance between the towers and cargo items may require different camera lenses with different focal lengths (eg, focal lengths). For example, a lens with a smaller focal length can be used in situations where cargo items pass closer to the sensor tower. When using a lens with a larger focal length, less information is captured closer to the sensor tower, but more accurate information is captured farther away from the tower.

In a particular embodiment, camera 206 is a wide-angle camera and the remaining cameras (202, 204, and 208-222) are high-precision cameras. In this embodiment, the wide-angle camera is positioned as camera 206 to provide sufficient height above the floor to allow the systems and methods described herein to better utilize the wide-angle field of view. The high-precision cameras 202, 204, and 208-222 capture higher-resolution images at different vertical viewpoints. The arrangement of high precision cameras 202, 204 and 208-222 provides overlapping fields of view, eg, as shown in "FIG. 5".

In some embodiments, cameras 202-222 are aimed perpendicular to the vertical axis of sensor tower 200. In other embodiments, one or more of the cameras 202-222 are aimed at different angles relative to the vertical axis of the sensor tower 200 to capture images from different angles and different viewing angles.

In some embodiments, the cameras 202-222 are synchronized such that all cameras 202-222 capture images at the same time. In certain embodiments, synchronization may be performed by triggering the control system of all other cameras in sensor tower 200 or the master camera. In some embodiments, cameras 202-222 have fast shutter speeds (eg, less than 100 microseconds).

The sensor tower 200 also includes four light bars 224, 226, 228, and 230 that illuminate cargo items passing through the dock. In some embodiments, the light bars 224-230 may flash at the same frequency as the images captured by the cameras 202-222. For example, the light bar can flash at 24Hz and the cameras 202-222 can capture images at 24Hz. Although four light bars 224-230 are shown, as shown in "FIG. 2," any number of light bars may be included in sensor tower 200 in other alternative embodiments. Additionally, other alternative embodiments may include lights having any shape, such as multiple round lights, square lights, and other shapes or configurations. In some embodiments, the strobing speed of the light bars 224-230 is three times faster than the camera captures the image. So if the camera is capturing images at 24 Hz, the light bars 224-230 are strobing at 72 Hz. In some embodiments, the brightness of one or more light bars 224-230 is adjusted based on ambient light levels. For example, if the ambient light level is low, one or more of the light bars 224-230 may be adjusted to a brighter level to better illuminate the cargo item. Similarly, if the ambient light level is high, one or more of the light bars 224-230 may be adjusted to a lower brightness level to avoid over-saturating the cargo item with light. In some embodiments, the brightness of one or more of the light bars 224-340 is adjusted to optimize detection and identification of labels, printing, markings, etc. associated with or associated with the item of cargo.

Sensor tower 200 also includes status lights 232 . In some embodiments, status light 232 is red when sensor tower 200 is not operating and green when sensor tower 200 is operating normally. The sensor tower 200 may not function, for example, when the dock door is closed or when the dock is blocked by items. In alternate embodiments, any number of status lights 232 may be included in sensor tower 200 . In these alternative embodiments, status lights 232 may be arranged in any configuration. For example, if sensor tower 200 has four status lights 232, the four lights may indicate network status, processing status, motion status, and general system status.

In the example of "FIG. 2," the sensor tower 200 includes various cameras 200-222 as sensors. In other embodiments, sensor tower 200 may include additional types of sensors, such as: radio frequency identification (RFID) sensors, light detection and ranging (light radar) sensors, thermal sensors, Time-of-Flight (ToF) sensors, proximity sensors, weight sensors, ultrasonic sensors, infrared sensors, and air purity sensors. In particular embodiments, sensor tower 200 may include any number of different types of sensors to capture different types of information. Data from all types of sensors can be aggregated to provide enhanced analysis of a cargo item and improved identification of one or more items associated with the cargo item.

In some embodiments, sensor tower 200 may include two sets of cameras 200 - 222 on opposite sides of sensor tower 200 . For example, a sensor tower 200 with two sets of cameras 200-222 may be positioned between two loading docks if they are close to each other. In some implementations, a first set of cameras 200-222 may capture a first set of images (associated with a first dock) processed by a first computing system, and a second set of cameras 200-222 may capture images processed by a first The second set of images (associated with the second dock) processed by the second computing system. This arrangement allows a single sensor tower 200 to capture images of cargo items on two adjacent docks.

"FIG. 3" illustrates an embodiment of a sensor tower 300 including fourteen cameras numbered 302-328. Sensor tower 300 may include at least one housing that provides support for cameras 302-328 and other components associated with sensor tower 300, as discussed herein. The housing can have any shape, any number of parts, and can be made of any one or more materials. Sensor towers 104 and 106 may be similar or identical to sensor tower 300 . In some embodiments, cameras 302-328 may be different types of cameras, such as wide-angle cameras and high-precision cameras. Cameras 302-328 may operate in the visible spectrum or any other spectrum. In some embodiments, the cameras 302-328 may have different lenses depending on the size (eg, width) of the dock.

In a particular embodiment, cameras 308 and 322 are wide-angle cameras and the remaining cameras (302-306, 310-320, and 324-328) are high-precision cameras. In this embodiment, the wide-angle cameras are positioned to maximize the field of view of the wide-angle cameras 308 and 322 within the sensor tower 300 operating range. The high-precision cameras 302-306, 310-320, and 324-328 capture higher-resolution images at different vertical viewpoints. The arrangement of high-precision cameras 302-306, 310-320, and 324-328 provides overlapping fields of view, eg, as shown in "FIG. 5".

In some embodiments, the cameras 302-328 are synchronized such that all cameras 302-328 capture images at the same time. In particular embodiments, synchronization may be performed using a control system or master camera that triggers all other cameras in sensor tower 300 . In some embodiments, cameras 302-328 have fast shutter speeds (eg, less than 100 microseconds).

The sensor tower 300 also includes four light bars 330, 332, 334, and 336 that illuminate cargo items passing through the dock. In some embodiments, the light bars 330-336 may flash at the same frequency as the images captured by the cameras 302-328. For example, the light bar can strobe at 24Hz and the cameras 302-328 can capture images at 24Hz. Although four light bars 330-336 are shown in FIG. 3, alternative embodiments may include any number of light bars in the sensor tower 300.

Sensor tower 300 also includes status lights 338 . In some embodiments, the status light 338 is red when the sensor tower 300 is not operating and green when the sensor tower 300 is operating normally. The sensor tower 300 may not function, for example, when the dock door is closed or when the dock is blocked by items.

In the example of "FIG. 3," a sensor tower 300 includes various cameras 302-328 as sensors. In other embodiments, the sensor tower 300 may include other types of sensors, such as: radio frequency identification (RFID) sensors, light detection and ranging (light radar) sensors, thermal sensors, Time-of-Flight (ToF) sensors, proximity sensors, weight sensors, and air purity sensors. In particular embodiments, sensor tower 300 may include any number of different types of sensors to capture different types of information. Data from all types of sensors can be aggregated to provide enhanced analysis of a cargo item and improved identification of one or more items associated with the cargo item.

“ FIG. 4 ” is a block diagram illustrating an embodiment of a sensor tower 400 . As shown in FIG. 4 , the sensor tower 400 includes a communication manager 402 , a processor 404 and a memory 406 . Communications manager 402 allows sensor tower 400 to communicate with other systems and components. Processor 404 executes various instructions to perform the functions provided by sensor tower 400, as discussed herein. Memory 406 stores these instructions and other data used by processor 404 and other modules and components included in sensor tower 400 .

Sensor tower 400 may also include graphics processing unit 408 , feedback system 410 and light manager 412 . Graphics processing unit 408 may process and manage various images captured by cameras associated with sensor tower 400. The graphics processing unit 408 can also manage the analysis of the captured images and the storage of the analysis results. In some embodiments, the graphics processing unit 408 is optimized for processing the types of images captured by the sensor towers discussed herein. For example, graphics processing unit 408 may analyze captured imagery to identify regions of interest within the imagery and identify specific markers or other identifiers that may determine the contents of an item associated with a cargo item. Feedback system 410 may provide feedback to users or systems located near sensor tower 400 . In some embodiments, feedback system 410 provides feedback in the form of optical signals, audio signals (eg, through a speaker), or the like. Light manager 412 may manage one or more light bars and one or more status lights associated with sensor tower 400 . For example, light manager 412 may determine whether the light bar should be activated and determine the color of the status light (eg, red or green).

Additionally, sensor tower 400 may include deep learning accelerator 414 , strobe controller 416 , Ethernet switch 418 , power distribution system 420 , camera 422 , and one or more storage devices 424 . The deep learning accelerator 414 can execute various deep learning systems, such as neural networks. In some embodiments, the deep learning accelerator 414 may identify one or more objects associated with the cargo item based on information detected in the captured imagery. Additional details regarding object recognition are discussed here. The strobe controller 416 can control the flickering frequency of the light bar. As discussed herein, in some embodiments, the light bar flashes at 24Hz to match the frequency of the camera. An Ethernet switch 418 may provide switching between sensor tower cameras and other systems or components. Power distribution system 420 may manage power distribution to various systems and components in sensor tower 400 . As discussed herein, camera 422 captures images of cargo moving through the dock. The one or more storage devices 424 can store various types of information, such as captured images, identified cargo items, identified objects associated with the cargo items, and the like. The storage device 424 includes, for example, a solid state disk, a memory device, and the like.

"FIG. 5" is an embodiment 500 of a cargo item moving across a dock between two sensor towers. As shown in "FIG. 5," sensor towers 502 and 504 are located on opposite sides of the dock. Cargo item 506 is passing through the dock. As discussed herein, sensor towers 502 and 504 each have a plurality of cameras that scan cargo item 506 as it passes through the dock. As shown in "FIG. 5", the wide-angle camera has a larger field of view 508 . In contrast, a high precision camera has a narrower field of view 510 . Multiple fields of view 508 and 510 overlap each other as shown in "FIG. 5".

"FIG. 6" illustrates an embodiment 600 of multiple docks, where each dock has two sensor towers 606 located on opposite sides of the dock. As shown in FIG. 6 , the first dock 602 includes a pair of sensor towers 606 , and the second dock 604 has another pair of sensor towers 606 . Although the example of "FIG. 6" shows only two docks 602 and 604, virtually any number of pairs of sensor towers 606 can be used with any number of docks. In some embodiments, a single sensor tower 606 may be used with each dock rather than a pair of sensor towers.

"FIG. 7" is a flowchart illustrating an embodiment of a process 700 for scanning a cargo item using two sensor towers as it moves through a dock. Initially, a cargo item approaches a dock (step 702). As the cargo item moves past the first sensor tower, the first sensor tower captures a plurality of images of a first side of the cargo item (step 704). As the cargo item moves past the second sensor tower, the second sensor tower captures a plurality of images of a second side of the cargo item (step 706).

Process 700 continues as the computing device analyzes the plurality of images of the first side of the cargo item and the second side of the cargo item (step 708). The computing device identifies at least one item associated with the cargo item (step 710). For example, an item may be an item on a pallet, an item in a container, or the like. In some embodiments, objects may be identified from captured images based on various attributes. For example, captured images may include product labels, barcodes, company logos, product logos, part numbers, product numbers, expiration dates, lot numbers, and other markings that identify the item and can associate the item with a particular manufacturer.

In some embodiments, some items associated with cargo items are not marked and do not have any other distinguishing markings. In these cases, the systems and methods described herein can determine dimensions and other characteristics of the item. For example, if the item is in a box, the systems and methods can determine the dimensions of the box, the color of the box, the reflectivity of the box, the texture of the box, the material used to make the box, and similar characteristics. In some embodiments, the item may be further identified based on the temperature of the item.

In some embodiments, the described systems and methods can also identify any damage to the item, evidence of tampering with the item, or evidence of leakage of the item's contents.

Referring again to process 700, the computing device further transmits information about the identified object to the operating platform (step 712). Process 700 continues as the operations platform updates various supply chain information and warehouse operations information based on the information about the identified items (step 714).

"FIG. 8" is a flow diagram illustrating an embodiment of a process 800 for scanning cargo items using a sensor tower as the cargo items move through the dock. Initially, a cargo item approaches a dock (step 802). As the cargo item moves past the sensor tower, the sensor tower captures a plurality of images of the cargo item (step 804). Process 800 continues as the computing device in the sensor tower analyzes the multiple images of the cargo item (step 806). The computing device then identifies at least one item associated with the cargo item (step 808). As mentioned above, the item may be an item on a pallet, an item in a container, or the like. Process 800 continues as the computing device transmits information about the identified object to the operating platform (step 810). The operation platform updates various supply chain information and warehouse operation information based on the information about the identified item (step 812).

"FIG. 9" illustrates an embodiment of a cargo record 900 for a particular cargo item scanned via the systems and methods discussed herein. The exemplary cargo record 900 includes information related to items that are delivered from a truck to a dock for a particular pallet. This information includes, for example, truck number, delivery time, pallet information, number of boxes, etc. Other embodiments of the item record 900 may include additional information not shown in "FIG. 9". In a specific implementation, a shipment record may include where the item came from or went to (e.g., address), carrier information, carrier identifier, pallet weight, special information about pallet contents (e.g., do not stack more than three multiple pallets or temperature restrictions), etc.

"FIG. 10" is a schematic diagram 1000 of an embodiment of a sensor tower. In the example in Figure 10, the sensor tower includes many of the components and systems discussed here. For example, schematic diagram 1000 illustrates Ethernet switch 1002 connected to camera array 1004, power strip 1006, and computing system (eg, edge computing unit) 1008. Also illustrated in the schematic is a gateway 1010 and a status indicator (eg, LED) 1012. In some embodiments, gateway 1010 is connected to Ethernet switch 1002 and switchboard 1006. The gateway 1010 can include wireless communication modules such as LTE (Long Term Evolution), WiFi, BLE (Bluetooth Low Energy) and GPS (Global Positioning System). Gateway 1010 may also include any number of antennas associated with wireless communication modules. One or more wireless communication modules can transmit data from the sensor tower to other systems, such as cloud-based computing devices.

"FIG. 10" further illustrates LED lighting 1014 and audio speakers 1016. The switchboard 1006 is also connected to a 24V AC/DC converter 1018 and a strobe controller 1020. The 24V AC/DC converter 1018 converts alternating current to 24V direct current. The 24V power rail is then connected to the power strip 1006 and other systems and components powered by the 24V. Although some systems and methods described herein use 24V, alternative embodiments may use different voltages. Additionally, some embodiments of sensor towers may use two or more voltages, eg: 5V, 12V, 24V, 48V, etc. The strobe controller 1020 manages the strobe frequency of the various lights associated with the sensor towers. The various components and systems shown in "Figure 10" are similar to those discussed herein.

"FIG. 11" illustrates a block diagram of an embodiment of an operating platform 1100. In some embodiments, operating platform 1100 is similar to operating platform 120 illustrated in FIG. 1 . As shown in FIG. 11 , the operating platform 1100 includes a sensor manager 1102 that manages various sensors (eg, cameras and other sensors) in one or more sensor towers. The hardware manager 1104 manages various features such as feedback lights, temperature control, and the like. For example, hardware manager 1104 may activate a status light for two seconds in response to detecting and scanning a cargo item. The operations platform 1100 also includes an informational texture detector 1106 that detects various textures associated with a cargo item or specific objects associated with a cargo item. The infotexture extractor 1108 can extract infotexture information from the infotexture data detected by the infotexture detector 1106 . The informative texture identifier 1110 identifies a particular cargo item or an object associated with a cargo item based on the detected and extracted informative texture data.

The operating platform 1100 also includes an edge-cloud communication manager 1112 that communicates with one or more cloud-based systems. For example, edge-cloud communication manager 1112 may send identified data to the cloud, send telemetry data to the cloud, receive operational commands from the cloud, and the like. The event handler 1114 can handle various events, such as truck arrival, truck departure, cargo loading event, cargo unloading event, cargo identification, item identification, and the like. The logistics data processor 1116 is capable of processing various types of logistics data, such as: processing high-level logistics data to infer information from previous steps (eg, truck inventory, truck load time, truck unload time, and dock throughput).

Additionally, the operating platform 1100 includes a monitoring and logging manager 1118 capable of monitoring and logging data associated with software and hardware operating status. For example, the monitoring and logging manager 1118 may log operating states associated with high CPU load, high temperature, sufficient memory or disk storage space, and the like. Data storage manager 1120 handles the storage and retrieval of data generated or used by any of the systems and components discussed herein. The data security manager 1122 is capable of protecting various types of data received and generated by the systems and components described herein. In some examples, the data security manager 1122 may handle data encryption, deletion of data upon detection of an intrusion, and the like.

The operating platform 1100 also includes an analysis module 1124 that performs various types of analysis to generate business insights and other data associated with the operation of the systems and methods described herein. Additionally included are dashboard and external application programming interface (API) modules 1126. For example, the dashboard may display various information related to movement of cargo items, movement of trucks, scheduling of shipments of cargo items, time delays, times to load or unload trucks or other vehicles, and the like. Machine learning module 1128 manages and performs various machine learning operations that may, for example, help identify cargo items, items associated with cargo items, and the like.

"FIG. 12" is a flowchart illustrating an embodiment of a process 1200 for managing operations associated with an operating platform. Initially, an operations platform (eg, operations platform 120 or 1100 ) detects that one or more items (eg, cargo items) are being loaded or unloaded by a loading dock (step 1202 ). In some embodiments, one or more items move proximate to one or more sensors, such as sensors contained in one or more sensor towers. The sensor tower captures a video feed (eg, a series of captured images) (step 1204).

The process 1200 continues with identifying and tracking systems and equipment operating in and near the dock (step 1206). In some embodiments, systems and devices may include forklifts, trucks, pallets, workers with wearable devices, individual items associated with cargo items, robotic devices, and the like. Process 1200 further detects informational textures associated with items being loaded or unloaded (step 1208). In some embodiments, the informational texture may include logos, labels, and other graphic indicia or graphic information of the types discussed herein. Process 1200 continues with extracting informative textures from video or images (step 1210).

In some embodiments, process 1200 may transmit the extracted informational texture data to the cloud (step 1212). This is optional, and in some embodiments the informational texture data is stored locally (eg, within a warehouse or other local facility). The extracted informational textures are then associated with a particular product or shipment (step 1214). For example, the system may associate an extracted informative texture with a particular product or shipment based on an informative texture identifier that identifies textures of known products or shipments.

The process 1200 continues with integrating specific product or cargo shipments into higher level logistics data (step 1216), such as inventory data, truck data, dock data, facility data, and the like. The higher level logistics data is then validated and compared for differences, and additional statistics are calculated (step 1218). For example, additional statistics may include truck inventory, truck load time, truck unload time, dock throughput, and the like.

"FIG. 13" is a flowchart illustrating an embodiment of a process 1300 for processing image data using a convolutional neural network (CNN). Initially, the process 1300 receives or captures wide-angle camera images (step 1302) and high-resolution camera images (step 1304). As discussed herein, wide-angle camera images 1302 and high-resolution camera images 1304 may be received (or captured) from any number of cameras in one or more sensor towers. Process 1300 continues with preprocessing all camera images (eg, all wide-angle camera images 1302 and all high-resolution camera images 1304 ) (step 1306 ). Examples of preprocessing operations might include scaling images, transforming images, and other image processing operations to change the original captured image into the format and content that the CNN expects as input.

As shown in "FIG. 13", the process 1300 may process wide-angle camera images with a first CNN (step 1308) and process high-resolution camera images with a second CNN (step 1310). In some embodiments, the described systems and methods can use any number of CNNs to process imagery and other types of data. In a specific implementation, the wide-angle camera image is a process of understanding the overall scene (for example, the overall loading and unloading dock information). High-precision camera images can be processed to identify information textures, extract information textures, etc. In some embodiments, the processing of the wide-angle camera imagery in step 1308 includes detection of objects such as forklifts, trucks, pallets, cargo items, and other items in loading docks. The processing (step 1310) of the high-resolution camera image can detect or recognize small object textures (eg, logos, plain text, and labels) and other specific information.

Process 1300 continues with receiving manual input on the CNN output (step 1312). In some embodiments, human input can verify the CNN's detection or recognition of the data. In other embodiments, human input can correct the CNN's detection or recognition of material. Process 1300 may further provide training feedback to the CNN based on human input (step 1314). As the CNN learns from human verification or correction, this training feedback may improve the CNN's future detection or recognition of material.

"FIG. 14" illustrates an example block diagram of a computing device 1400 suitable for implementing the systems and methods described herein. In some embodiments, a cluster of computing devices interconnected by a network may be used to implement any one or more components of the systems discussed herein.

Computing device 1400 may be used to perform various processes, such as those discussed herein. Computing device 1400 may function as a server, client, or any other computing entity. A computing device may perform various functions discussed herein, and may execute one or more applications, such as the applications described herein. The computing device 1400 may be any one of various computing devices, such as a desktop computer, a notebook computer, a server computer, a handheld computer, a tablet computer, and the like.

Computing device 1400 includes one or more processors 1402, one or more memory devices 1404, one or more interfaces 1406, one or more mass storage devices 1408, one or more input/output (I/O) Device 1410 and display device 1430 are all coupled to bus bar 1412 . Processor 1402 includes one or more processors or controllers that execute instructions stored in memory device 1404 and/or mass storage device 1408 . The processor 1402 may also include various types of computer-readable media, such as cache memory.

Memory device 1404 includes various computer-readable media such as volatile memory (e.g., random access memory (RAM) 1414) and/or nonvolatile memory (e.g., read only memory (ROM) 1416) . The memory device 1404 may also include rewritable ROM, such as flash memory.

The mass storage device 1408 includes various computer-readable media, such as magnetic tapes, magnetic disks, optical disks, solid-state memory (eg, flash memory), and the like. As shown in FIG. 14 , the specific mass storage device is a hard disk 1424 . Various drives may also be included in mass storage device 1408 to enable reading and/or writing to various computer-readable media. Mass storage device 1408 includes removable media 1426 and/or non-removable media.

Input/output (I/O) devices 1410 include various devices that allow data and/or other information to be input to or retrieved from computing device 1400 . Exemplary input/output (I/O) devices 1410 include cursor control devices, keyboards, keypads, microphones, monitors or other display devices, speakers, printers, network interface cards, modems, cameras, CCDs or other Image capture equipment, etc.

Display device 1430 includes any type of device capable of displaying information to one or more users of computing device 1400 . Examples of the display device 1430 include a monitor, a display terminal, a video projection device, and the like.

Interfaces 1406 include various interfaces that allow computing device 1400 to interact with other systems, devices, or computing environments. Exemplary interfaces 1406 include any number of different network interfaces 1420, such as local area network (LAN), wide area network (WAN), wireless network, and Internet interfaces. Other interfaces include a user interface 1418 and a peripherals interface 1422 . Interface 1406 may also include one or more user interface elements 1418 . Interfaces 1406 may also include one or more peripheral interfaces, such as interfaces for printers, pointing devices (mouse, touchpad, etc.), keyboards, and the like.

Bus 1412 allows processor 1402 , storage device 1404 , interface 1406 , mass storage device 1408 , and input/output (I/O) device 1410 to communicate with each other, as well as other devices or components coupled to bus 1412 . Buses 1412 represent one or more of several types of bus structures, such as a system bus, a PCI bus, an IEEE 1394 bus, a USB bus, and the like.

For purposes of illustration, programs and other executable program components are presented herein as separate functional blocks, although it is understood that such programs and components may reside in different storage components of computing device 1400 at different times and be executed by processing implementor(s) 1402. Alternatively, the systems and programs described herein may be implemented in hardware or a combination of hardware, software, and/or firmware. For example, one or more application specific integrated circuits (ASICs) may be programmed to implement one or more of the systems and procedures described herein.

While various embodiments of the present disclosure have been described herein, it should be understood that they have been presented by way of example only, and not limitation. It will be apparent to those skilled in the relevant art that various changes in form and details can be made therein without departing from the spirit and scope of the present disclosure. Thus, the breadth and scope of the present disclosure should not be limited by any of the described exemplary embodiments, but should be defined only in accordance with the appended claims and their equivalents. The description herein has been presented for purposes of illustration and description. It is not intended to be exhaustive or to limit the disclosure to the precise form disclosed. Many modifications and variations are possible in light of the disclosed teaching. Furthermore, it should be noted that any or all of the alternative embodiments discussed herein may be used in any combination desired to form additional hybrid embodiments of the present disclosure.

100: Environment 102: loading and unloading table 104: First sensor tower 106:Second sensor tower 108: cargo items 110: Data communication network 112:Wearable devices 114:Robot equipment 116: forklift 118: Warehouse equipment 120: Operating platform 122: Cloud-Based Computing Systems 200: Sensor Tower 202, 204, 206, 208, 210, 212, 214, 216, 218, 220, 222: camera 224, 226, 228, 230: Light bar 232: status light 300: Sensor Tower 302, 304, 306, 308, 310, 312, 314, 316, 318, 320, 322, 324, 326, 328: camera 330, 332, 334, 336: Light bar 338: status light 400: Sensor Tower 402: Communication Manager 404: Processor 406: Memory 408: Graphics Processing Unit 410: Feedback Systems 412: Light Manager 414:Deep Learning Accelerator 416: Strobe controller 418:Ethernet switch 420: Distribution system 422: camera 424: storage device 500: Example 502, 504: sensor tower 506: cargo items 508, 510: Vision 600: Example 602: The first loading and unloading platform 604: Second loading and unloading platform 606:Sensor Tower 700: process 800: process 900: cargo record 1000: Schematic diagram 1002:Ethernet switch 1004: camera array 1006: Switchboard 1008:Computing system 1010: Gateway 1012: status indicator 1014: LED lighting 1016: Audio speaker 1018:24V AC/DC Converter 1020: Strobe controller 1100: Operating platform 1102: Sensor Manager 1104: hardware manager 1106: Information Texture Detector 1108: Info Texture Extractor 1110: Information Texture Recognizer 1112:Edge-Cloud Communication Manager 1114:Event handler 1116:Logistics Data Processor 1118:Monitoring and Recording Manager 1120: Data storage manager 1122:Data Security Manager 1124: Analysis module 1126: Dashboard and External Application Programming Interface (API) Modules 1128:Machine Learning Module 1200: process 1300: process 1400: computing equipment 1402: Processor 1404: memory device 1406: interface 1408: mass storage device 1410: Input/Output (I/O) Devices 1412: busbar 1414: random access memory 1416: ROM 1418: user interface 1420: Network interface 1422: Peripheral device interface 1424:hard disk 1426: Removable media 1430: display device Step 702: The cargo item approaches the dock Step 704: The first sensor tower captures a plurality of images of the first side of the cargo item as the cargo item moves past the first sensor tower Step 706: The second sensor tower captures a plurality of images of a second side of the cargo item as the cargo item moves past the second sensor tower Step 708: The computing device analyzes the plurality of images of the first side of the cargo item and the second side of the cargo item Step 710: The computing device identifies at least one object associated with the cargo item Step 712: The computing device further transmits information about the identified object to the operating platform Step 714: The operation platform updates various supply chain information and warehouse operation information based on the information about the identified items Step 802: The cargo item approaches the dock Step 804: The sensor tower captures multiple images of the cargo item as the cargo item moves past the sensor tower Step 806: The computing device in the sensor tower analyzes the multiple images of the cargo item Step 808: The computing device identifies at least one item associated with the cargo item Step 810: The computing device transmits information about the identified object to the operating platform Step 812: The operation platform updates various supply chain information and warehouse operation information based on the information about the identified object Step 1202: The operating platform detects that one or more items are being loaded or unloaded by the dock Step 1204: The sensor tower captures a video feed (e.g., a series of captured images) Step 1206: Identify and track systems and equipment operating in and near the dock Step 1208: Detect informational textures associated with items being loaded or unloaded Step 1210: Extract information texture from video or image Step 1212: Send the extracted information texture data to the cloud Step 1214: Associating the extracted informative texture with a specific product or shipment Step 1216: Consolidate specific product or cargo shipments into higher level logistics data (e.g. inventory data, truck data, dock data, facility data) Step 1218: Verify higher level logistics data and compare differences, and calculate additional statistics Step 1302: Receive or retrieve wide-angle camera images Step 1304: Receive or capture high-precision camera images Step 1306: Preprocess all camera images Step 1308: Process the wide-angle camera image with the first CNN Step 1310: Process the high-precision camera image with the second CNN Step 1312: Receive manual input on CNN output Step 1314: Provide training feedback to CNN based on human input

Figure 1 illustrates a block diagram of an environment in which exemplary embodiments may be implemented. Figure 2 illustrates an embodiment of a sensor tower including eleven cameras. Figure 3 illustrates an embodiment of a sensor tower including fourteen cameras. Figure 4 illustrates a block diagram of an embodiment of a sensor tower. Figure 5 illustrates an embodiment where cargo passes through a dock between two sensor towers. Figure 6 illustrates an embodiment of multiple docks, where each dock has two sensor towers located on opposite sides of the dock. Figure 7 illustrates an embodiment flow diagram of a process for scanning cargo items using two sensor towers as they pass through a dock. Figure 8 illustrates a flowchart of an embodiment of a process for scanning cargo items using a sensor tower as they pass through a dock. Figure 9 illustrates an embodiment of a cargo record for a particular cargo item scanned using the systems and methods discussed herein. Figure 10 illustrates a schematic diagram of one embodiment of a sensor tower. Figure 11 illustrates a block diagram of an embodiment of an operating platform. Figure 12 illustrates a flow diagram of an embodiment for managing operational processes associated with an operating platform. FIG. 13 illustrates a flowchart of an embodiment of a process of processing image data using a convolutional neural network. Figure 14 illustrates an exemplary block diagram of a computing device.

100: Environment

102: loading and unloading table

104: First sensor tower

106:Second sensor tower

108: cargo items

110: Data communication network

112:Wearable devices

114:Robot equipment

116: forklift

118: Warehouse equipment

120: Operating platform

122: Cloud-Based Computing Systems

Claims (20)

A cargo management system comprising: one or more processors; and One or more non-transitory computer-readable media storing instructions executable by the one or more processors, wherein the instructions, when executed, cause the system to perform operations including: receiving at least one wide-angle camera image from a sensor tower, wherein the sensor tower is located adjacent to a dock and the wide-angle camera image is associated with at least a portion of the dock; receiving a plurality of high-precision camera images from the sensor tower, wherein the plurality of high-precision camera images are associated with at least a portion of the dock; preprocessing the wide-angle camera image and the plurality of high-resolution camera images to generate a pre-processed wide-angle image and a plurality of pre-processed high-resolution images; processing the preprocessed wide-angle image using a first convolutional neural network (CNN); and The plurality of pre-processed high-resolution images are processed using a second convolutional neural network (CNN). The cargo management system as claimed in claim 1, further comprising identifying at least one device operating near the loading dock based on processing the plurality of pre-processed high-resolution images using the second CNN. The cargo management system of claim 1, the operations further comprising identifying a cargo item approaching the dock based on processing the plurality of pre-processed high-resolution images using the second CNN. The cargo management system as recited in claim 3, further comprising identifying a plurality of objects associated with the cargo item based on processing the plurality of pre-processed high-resolution images using the second CNN. The cargo management system as described in Claim 3, wherein identifying the cargo item includes analyzing the label on the cargo item, analyzing the text on the cargo item, analyzing the logo on the cargo item, analyzing the size of the cargo item, analyzing the At least one of the color of the cargo item or analyzing the volume of the cargo item. In the cargo management system as claimed in claim 1, the operation further includes receiving a manual input based on the result of processing the plurality of pre-processed high-resolution images using the second CNN. As the goods management system described in claim 6, the operation further includes training the second CNN based on the received manual input. The cargo management system of claim 1, the operations further comprising analyzing the pre-processed high-resolution images to identify whether at least one object associated with the cargo item is damaged or to identify at least one event associated with the cargo item whether it has been tampered with. The cargo management system of claim 1, the operation further comprising determining whether the cargo item is being loaded or unloaded by the dock. The cargo management system as recited in claim 1, the operations further comprising identifying at least one device operating in the vicinity of said dock. The cargo management system of claim 1, the operations further comprising detecting an information texture associated with at least one item near the dock based on processing the plurality of pre-processed high-resolution images using the second CNN. The cargo management system of claim 11, the operations further comprising transmitting the information texture detected to be associated with at least one object proximate to the dock to a remote computing system. In the item management system of claim 11, the operation further includes associating the information texture with a specific product. In the cargo management system as claimed in claim 13, the operation further includes integrating the specific product data with higher-level logistics data. A cargo management method, comprising the following steps: receiving at least one wide-angle camera image from a sensor tower, wherein the sensor tower is located adjacent to a dock and the at least one wide-angle camera image is associated with at least a portion of the dock; receiving a plurality of high-precision camera images from the sensor tower, wherein the plurality of high-precision camera images are associated with at least a portion of the dock; processing the wide-angle image using a first convolutional neural network (CNN); processing the plurality of high-resolution images using a second convolutional neural network (CNN); and A cargo item near the loading and unloading dock is identified according to the processed plurality of high-precision images. The cargo management method according to claim 15, further comprising preprocessing the at least one wide-angle camera image and the plurality of high-resolution camera images to generate a pre-processed wide-angle image and a plurality of pre-processed high-resolution images. The cargo management method as described in claim 15, further comprising identifying a plurality of objects associated with the cargo item based on using the second CNN to process the plurality of high-precision images. The cargo management method as described in claim 15, wherein identifying the cargo item includes analyzing the label on the cargo item, analyzing the text on the cargo item, analyzing the mark on the cargo item, analyzing the size of the cargo item, analyzing the At least one of the color of the cargo item or the volume of the cargo item analyzed. The cargo management method according to claim 15, further comprising analyzing the processed plurality of high-precision images to identify whether at least one item associated with the cargo item is damaged or to identify at least one item associated with the cargo whether it has been tampered with. The cargo management method according to claim 15, further comprising training the second CNN based on receiving a manual input associated with the processed high-resolution images.

Images