TW202326634A - Intelligent railway monitoring system and method thereof - Google Patents

Abstract

An intelligent railway monitoring system includes: a cloud database module for storing the image data of driving rail vehicle; a central monitoring system connects to the cloud database module for data signal transmission; an on-board subsystem of the driving rail vehicle includes an image capturing module, a real-time image storage module, a data transmission module, an image reference module, an image recognition module, warning alert module and a control module can link to both cloud database module and central monitoring system. During the driving of the rail vehicle, the image capturing module can generate a set of real-time images to compare with the images from the image reference module. The difference of both sets of images can be retrieved and recorded in the computer system. When the image recognition module recognizes that there is an abnormal state between the image to be compared and the reference image, the warning alert module will send a warning signal to the controller.

Description

Intelligent track monitoring system and method thereof

The present invention relates to a smart track monitoring system and its method, especially a smart track vehicle application, by capturing the driving record images of the track vehicles, and through image recognition, comparing the image screen in front of the track vehicles to see whether there is an abnormal state of intelligence Track monitoring system and method thereof.

At present, the safety of rail vehicles, such as high-speed railways, general railways, light rail trains and rapid transit, depends on the vigilance and attention of the driver, or the traffic control center in the station understands the status of the track through the monitor screen In addition, in most cases, after an accident occurs, it is not known which section of the track has an accident or abnormality after the notification. Therefore, it is difficult to achieve timely risk prevention work.

The known technology of the Republic of China Patent No. I318187 is mainly to provide images for rail vehicles to watch, increase the driver's field of vision, and reduce the manpower expenditure of rail patrols. The known technology of the Republic of China Patent No. I307325 is to install a micro-motion sensor on the track and use the difference in vibration frequency to judge whether the track itself is damaged, loose or other abnormal conditions. Known technology The Republic of China Announcement Patent No. I304191 carries out the judgment of the road condition ahead of the track through an unmanned leading vehicle, and sends the signal back on the rail car.

However, the technology published by the Republic of China Patent No. I318187 still needs to be judged by the driver's naked eyesight. The conventional technology published by the Republic of China Patent No. I307325 cannot know whether there is any foreign object on the track. The conventional technology published by the Republic of China Patent No. I304191 requires another Judging by an unmanned leading vehicle, its construction is complicated, the cost is too high, and it reveals the status of identifying track abnormalities through image recognition. To sum up, the above-mentioned conventional technology still does not achieve automatic judgment by the driving system of whether there is an abnormality or an abnormal state on the track ahead and an early warning mechanism. Therefore, it needs to be improved. The present invention uses the image recognition system installed on the rail vehicle itself to automatically identify and determine the status of the track ahead of the track and whether there is any abnormality around the track at any time, which can solve the shortcomings of the conventional technology.

An object of the present invention is to provide a smart track monitoring system and its method, which has the function of pre-judging whether there is any abnormality and what kind of abnormal condition in a certain section of the track ahead through image recognition.

Another object of the present invention is to provide an intelligent track monitoring system and its method, which has the function of immediately notifying the occurrence of abnormal conditions in a certain track section of adjacent track vehicles.

In order to achieve the above and other objectives, the intelligent track monitoring system of the present invention includes: a cloud database module, which stores the driving record image data of rail vehicles; a central monitoring main system, which is electrically connected to the cloud database module, and performs data Signal transmission; and at least one vehicle-mounted subsystem, each of which is installed on a plurality of rail vehicles and electrically connected to the cloud database module and the central monitoring main system, the vehicle-mounted subsystem includes: an image capture A module for recording rail vehicle driving images, generating a real-time driving record image, and capturing the real-time driving record image to generate an image for comparison; a real-time image storage module for storing the real-time driving record image and the desired comparison image Image; a data transmission module, uploading the real-time driving record image to the cloud database module and downloading a reference image from the cloud database module; a reference image storage module, storing the reference image; an image recognition A module for identifying the difference between the image to be compared and the reference image; an abnormality warning module, when the image recognition module recognizes that there is an abnormal state between the image to be compared and the reference image, the abnormality warning module a warning signal; and a control module electrically connected to the image capture module, the real-time image storage module, the reference image storage module, the data transmission module, the image recognition module and the An abnormal warning module, and controls the operation mechanism among the modules, wherein, the image information of the image to be compared and the reference image includes recording date and time and latitude and longitude coordinates.

In order to achieve the above and other purposes, the intelligent track monitoring method of the present invention includes: downloading a reference image from a cloud database module by a data transmission module, storing the reference image in a reference image storage module; The image capture module obtains a real-time driving record image, and then captures an image to be compared from the real-time driving record image, and stores the image to be compared in a real-time image storage module; the data transmission module will The real-time driving record image is sent to the cloud database module for storage; a reference latitude and longitude coordinate and a latitude and longitude coordinate to be compared are respectively extracted from the reference image and the image to be compared, and the reference latitude and longitude coordinate and the to be compared image are used Image overlap correction is performed on the latitude and longitude coordinates, so that the image comparison reference point of the reference image and the image to be compared is within the error range of the coordinates, and the image difference is performed on the frame of the image to be compared and the reference image according to the coordinate position of the image Comparison: When performing image difference comparison, if there is any abnormality in the image to be compared after identification and comparison, an abnormality warning module installed on the rail vehicle will send out an abnormality warning signal.

In some embodiments of the present invention, the image recognition module further includes a distance measuring unit for calculating the distance between the obstacle ahead and the rail vehicle.

In some embodiments of the present invention, the abnormal warning module further includes an instant notification unit, which transmits the warning signal to the central monitoring main system or other rail vehicles.

In some embodiments of the present invention, the control module further includes a braking unit electrically connected to the braking system of the rail vehicle to reduce the speed of the rail vehicle.

In some embodiments of the present invention, there is a comparison period when the images to be compared and the frames of the reference image are compared for image differences.

In some embodiments of the present invention, when performing image difference comparison, if there is an abnormality in the picture of the image to be compared after identification and comparison, an abnormal range is defined in the image to be compared, according to the current state of the rail vehicle. The driving speed and latitude and longitude coordinates calculate how long it takes to reach the coordinate position of the different range within a unit of time.

In some embodiments of the present invention, after an abnormal range is defined in the image to be compared, it is judged whether the abnormal range is on the moving track, on the opposite track or across the two-way track.

In some embodiments of the present invention, when image difference comparison is performed, a comparison range is respectively defined on the frame of the image to be compared and the frame of the reference image, and after identification and comparison, if the frame of the image to be compared has abnormal point, but the position of the abnormal point falls outside the comparison range, it is still judged as no abnormal state.

In some embodiments of the present invention, when image difference comparison is performed, if there is an abnormality in the image to be compared after identification and comparison, the abnormality warning module activates a braking unit, and the braking unit reduces the speed of the rail vehicle. speed.

FIG. 1 is a structure diagram of an embodiment of the smart track monitoring system of the present invention, please refer to FIG. 1 . The intelligent track monitoring system of the present invention comprises: a cloud database module (10), a central monitoring module (20) and at least one on-board subsystem (30), and the cloud database module (10) stores rail vehicles Driving record image data, the central monitoring main system (20) is electrically connected to the cloud database module (10) for data signal transmission, each of the on-board subsystems (30) is respectively set on a plurality of rail vehicles (RV), and The cloud database module (10) and the central monitoring main system (20) are respectively electrically connected.

Each of the on-board subsystems (30) preferably has a body identification code, as the identity identification of the rail vehicle provided by it, the on-board subsystem (30) includes: an image capture module (31), a real-time image storage Module (32), a data transmission module (33), a reference image storage module (34), an image recognition module (35), an abnormal warning module (36) and a control module (37) , wherein, the image capture module (31) is preferably a high-magnification telephoto camera for recording rail vehicle driving images, generating a real-time driving record image, and performing frame capture to generate a desired comparison image (IC).

The image capture module (31) has a satellite positioning unit. When the track vehicle driving record is executed, the satellite positioning unit will simultaneously record the time and latitude and longitude coordinates in the image file data of the driving record. Therefore, the desired comparison Each frame of image (IC) (image frame per unit time) has time series and latitude and longitude coordinate information at the same time, and the image frame at a specific time point or specific coordinate can be locked through time series or latitude and longitude coordinates. In this embodiment, according to actual needs, an image capture cycle is set to capture image frames to generate the image to be compared (IC), for example, once every 5 seconds, so that the image to be compared (IC) has time and coordinate information as a reference for image comparison. In addition, the image to be compared (IC) can also be based on the driving distance, for example, an image is captured every 10 meters, and the driving distance can be calculated by the straight-line distance between two points or the actual driving distance.

The real-time image storage module (32) stores the real-time driving record image and the image to be compared (IC). The real-time driving record image can be used as a reference image for subsequent image comparison of other rail vehicles. After the image capture module (31) captures the image to be compared (IC), it is sequentially stored in the real-time The image storage module (32) is used as an image frame comparison to determine whether there is an abnormal condition on the driving track.

The data transmission module (33) uploads the real-time driving record image to the cloud database module (10) and downloads a reference image (IR) from the cloud database module (10). The real-time driving record image is uploaded to the cloud database module (10) for storage and backup, which can be used as a source of subsequent reference images, and the reference image (IR) is used for image comparison, and is stored in the A reference image storage module (34), wherein the reference image (IR) also includes information such as time series and latitude and longitude coordinates.

The image identification module (35) identifies the difference between the image to be compared (IC) and the reference image (IR). Since the track on which the rail vehicle travels and its surrounding images are fixed in principle, there will not be much change, such as: high-speed rail track, Taiwan Railway track, MRT track or light rail MRT track, using high-resolution driving records The image screen can push the recognition range forward to hundreds of meters, or even farther, and then calculate the relative position or distance between a certain point or range in the pixel of the image screen and the rail vehicle. Under normal circumstances, under the same coordinate position, the image frames of the image to be compared (IC) and the reference image (IR) will not have abnormal changes except for weather factors or differences in brightness during the day and night. , such as the track itself and its surroundings, unless the track is deformed or there are large foreign objects on the track, or other track vehicles, such as engineering vehicles.

The image recognition module (35) selects the comparison coordinate points according to the coordinate positions in the image to be compared (IC) and the reference image (IR) to compare the two frames, and compares the two images through image recognition. Whether there is a difference in the image screen, and find out the position of the difference point in the image to be compared (IC) screen. For example, when the rail vehicle comes to a certain coordinate position, that is, a certain section of the track, after the coordinate position is determined, under normal circumstances, the image capture module (31) captures the image to be compared ( IC) and the frame of the reference image (IR) overlap, and the content of the two frames is consistent, which means that there is no abnormal situation on the track path ahead.

FIG. 4 is a schematic diagram of image comparison of an embodiment of the smart track monitoring method of the present invention, please refer to FIG. 4 . The image recognition module (35) selects the comparison coordinate points according to the coordinate positions in the image to be compared (IC) and the reference image (IR) to compare the two images, if the difference is unimportant or compared The difference outside the range, for example, the difference between the image to be compared and the reference image is at a certain position in the image frame, but the actual distance of the difference is far from the track, which does not affect driving safety. Therefore, it is still judged that the two screens are the same, and there is no abnormal situation.

FIG. 2 is a structure diagram of another embodiment of the smart track monitoring system of the present invention, please refer to FIG. 2 . Preferably, the image recognition module (35) further includes a distance measuring unit (350), if there is an abnormality in the picture of the comparison image (IC), by calculating the distance between a certain point or a certain area in the image picture and the range of the rail vehicle The relative position or distance can be calculated according to the focal length of the lens of the image capture module (31), the size of the pixel (Pixel) and the ratio of the obstacle (subject), and the position and position of the obstacle or abnormal shape in front can be calculated. Distance L between rail vehicles, where, distance L = focal length , the subject is a foreign object or obstacle, the focal length and pixel size are known, and the subject size (length or width) can be determined through image recognition to determine what kind of object the subject belongs to. For example, first determine whether it is a passenger car or a truck Or track engineering vehicles, and the size (length or width) of these types of vehicles can be calculated in advance by setting several sets of preset values according to the types of vehicles. Or, since the width of the track is a fixed size, the actual size of the subject can be calculated based on the ratio of the length and width of the foreign object or obstacle in the image to the width of the track, as calculation data, such as some falling objects, such as large rocks , scaffolding and the like. Through the function of the distance measuring unit (350), relative response measures can be taken according to the vehicle speed at that time, such as deceleration, for emergency treatment.

Fig. 3 is a schematic diagram of a foreign object obstructing the track according to an embodiment of the intelligent track monitoring system of the present invention. When the image recognition module (35) recognizes that there is an abnormal state between the image to be compared and the reference image, the abnormality warning module (36) sends out a warning signal. The warning signal can be sound, light or electronic signal to notify the rail vehicle driver.

Please continue to refer to FIG. 2 and FIG. 3 . Preferably, the abnormal warning module (36) further includes an instant notification unit (360), which transmits the warning signal to the central monitoring main system (20) or other rail vehicles. The central monitoring main system (20) is the dispatching and monitoring center of all rail vehicles. When receiving the warning signal, it can immediately carry out emergency treatment, for example; notify the rail vehicle that is about to pass by the abnormal situation through the identification code to improve Be vigilant, pay attention to the road conditions ahead, or directly notify other rail vehicles that are about to pass through the road section by the rail vehicle that finds an abnormal situation to raise their vigilance.

The control module (37) is electrically connected to the image capture module (31), the real-time image storage module (32), the reference image storage module (33), the data transmission module (34), The image recognition module (35) and the abnormal warning module (36) control the operation mechanism between the modules, and are connected to the central monitoring main system (20) network through the control module (37) Lines for the transmission of control signals and data data.

Please continue to refer to Figure 2. Preferably, the control module (37) further includes a braking unit (370), which is electrically connected to the braking system of the rail vehicle (RV) to reduce the speed of the rail vehicle (RV). When the abnormal warning module (36) sends out the warning signal, it can automatically start the deceleration function according to the setting of the braking unit (370), so that the rail vehicle (RV) can automatically decelerate adaptively according to the current vehicle speed, For example; when the image recognition module (35) finds that there are abnormalities or foreign objects on the track ahead, such as general vehicles stuck on the track, or when the track engineering vehicle performs operations, but forget to notify, at this time, the comparison image ( IC) will be identified and compared by the image recognition module (35) to be abnormal, because there is no vehicle on the track in the road section in the reference image (IR). Therefore, the braking unit (370) can immediately and automatically perform appropriate deceleration actions according to the current vehicle speed adaptively, so as to ensure that the rail vehicle (RV) will not derail or overturn due to sudden braking or hitting a foreign object in front, which can reduce casualties minimized.

The intelligent track monitoring method of the present invention includes: downloading a reference image from a cloud database module by a data transmission module, and storing the reference image in a reference image storage module for use as a reference comparison image, wherein , the reference image storage module is set on a rail vehicle.

An image to be compared is obtained from an image capture module, wherein the image capture module is set on the rail vehicle, and the image to be compared is stored in a real-time image storage module. The image capture module is preferably arranged in front of the front of the rail vehicle, records the front road condition image when the rail vehicle is moving, generates a real-time driving record image, and stores the real-time driving record image in the real-time image storage module, the The image capture module performs frame capture from the real-time driving record image to generate an image to be compared, wherein the real-time driving record image includes the track and its surrounding images, and includes time series and latitude and longitude coordinate information.

Preferably, there is a comparison period when the images to be compared and the frames of the reference image are compared for image differences. The comparison cycle can be unit time or unit distance. The image to be compared is extracted from the real-time driving record image according to actual needs, and a cycle or driving distance is set for image acquisition, for example, every 5 seconds or Capture images once every 10 seconds or every 10 meters or 20 meters. You can adjust the capture method, time period or driving distance according to the environment of the driving section, and have obtained better images for comparison. Among them, the obtained The plural images to be compared also include time and latitude and longitude coordinate information.

The data transmission module transmits the real-time driving record image to the cloud database module for storage. The real-time driving record image is uploaded to the cloud database module for storage and backup, which can be used as a source of subsequent reference images, that is, the real-time driving record image recorded by the earlier rail vehicle is used as the subsequent later rail vehicle Therefore, the reference image also includes information such as time series and latitude and longitude coordinates.

Then, using an image recognition module to extract a reference latitude and longitude coordinate and a latitude and longitude coordinate information to be compared from the reference image and the image to be compared respectively, and performing image overlap correction by using the reference latitude and longitude coordinate and the latitude and longitude coordinate to be compared Make the frame comparison reference point of the reference image and the image to be compared within the coordinate error range, and perform image difference comparison between the frames of the image to be compared and the reference image according to the coordinate positions of the images.

Since the coordinates of each image frame will affect the actual latitude and longitude coordinates due to factors such as the GPS positioning system itself, vehicle speed or other interference, so when the frame of the reference image is compared with the frame of the image to be compared , it is better to perform coordinate correction first. For example, if the coordinate error range is 10 meters, it means that the error range within 10 meters before and after the coordinate point is within a reasonable error range. In this embodiment, the coordinate error range is 5 meters, that is, the image recognition module will, in the process of recognition and comparison, take the distance of 5 meters before and after the comparison coordinate point between the image to be compared and the reference image The images are compared sequentially. If there is at least one identical image between the image to be compared and the reference image after the comparison within the coordinate error range, it means that there is no abnormality; if there is no identical image in the entire comparison range, it means There are abnormalities. In addition, in order to avoid misjudgment, the determination of the number of occurrences of the same screen can be increased. In this embodiment, it is set to compare the appearance of the same screen twice to indicate that there is no abnormal situation.

Please continue to refer to FIG. 4 . When performing image difference comparison, a comparison range is defined on the frame of the image to be compared and the frame of the reference image. If it is outside the comparison range, it is still judged to be in a non-abnormal state. The determination of the same picture can be based on the actual needs of image recognition and ignore the differences that are not important or outside the scope of the comparison. For example, the difference between the image to be compared and the reference image is in a certain part of the image frame, but the difference If it falls outside the comparison range, it means that the difference is far from the actual distance and the track, which does not affect driving safety. Therefore, it is still determined that there is no abnormality.

FIG. 5 is a schematic diagram of an abnormal state of an embodiment of the smart track monitoring method of the present invention, please refer to FIG. 5 . When performing image difference comparison, if there is any abnormality in the image to be compared after identification and comparison, an abnormality warning module installed on the rail vehicle will send out an abnormality warning signal. When there is a discrepancy between the frame of the image to be compared and the frame of the reference image, the abnormality warning module sends out the abnormality warning signal to inform the driver to pay attention. The warning signal can be sound, light or electronic signal.

Preferably, when image difference comparison is performed, if there is an abnormality in the image to be compared after identification and comparison, the abnormality warning signal is sent to a central monitoring main system or other rail vehicles. When an abnormal situation occurs, the abnormal warning signal not only informs the driver of the rail vehicle, the abnormal warning module also transmits the electronic signal of the abnormal warning signal to the central monitoring main system or other preparations to pass through the abnormal point or accident point The central monitoring main system will make emergency notifications based on the route and location of each rail vehicle, or the rail train will directly notify other rail vehicles that are about to pass the abnormal point or accident point, so that other rail vehicles can prepare , to avoid serial accidents.

FIG. 6 is a schematic diagram of an abnormal state of another embodiment of the smart track monitoring method of the present invention. Please continue to refer to FIG. 6 and continue to refer to FIG. 3 . Preferably, when performing image difference comparison, if there is an abnormality in the picture of the image to be compared after identification and comparison, an abnormal range is defined in the abnormal area in the image to be compared, according to the current driving of the rail vehicle The speed and latitude and longitude coordinates calculate how long it takes to reach the coordinate position of the range within a unit of time. When there is a foreign object on the image to be compared, the abnormal range data is obtained after defining the range of the foreign object, and then the actual distance is calculated according to the pixel (Pixel) ratio based on the position in the image screen, and then the current latitude and longitude coordinates of the rail vehicle And the coordinate position within the range within which the vehicle speed is calculated can provide drivers with corresponding measures for emergency situations. The distance between the rail vehicle and the foreign body L=focal length , the subject is a foreign object or an obstacle, the focal length and pixel size are known, and the subject size (length or width) can be obtained by calculating the ratio of the abnormal range data to the angiographic image. In addition, in order to more accurately calculate how long it takes to reach the coordinates of the foreign object T, the time from the comparison to the abnormality and the calculation of the distance can be calculated , according to the forward speed V of the rail car to calculate the forward distance during this period ,in, V , then T= .

Please continue to refer to FIG. 3 and FIG. 5 . Preferably, after an abnormal range is defined in the image to be compared, it is determined whether the abnormal range is on the moving track, on the opposite track or across the two-way track according to the position area of the abnormal range in the image to be compared. The track used by the general rail vehicle is usually close to the track in the two traveling directions, and only the safe distance between the two opposite rail vehicle bodies is maintained. Therefore, the location of the abnormal point or abnormal range in the image to be compared can be compared range, to deduce whether the abnormal point is on the track, on the opposite track or across the two-way track. For example, if the range of the abnormal point is within the width of the track, it will not affect the opposite lane. For example, if a small unknown object falls If the abnormal point is wider than the width of the track, it means that the foreign object may span two lanes and may affect the opposite lane. For example, a large truck breaks down on the track. At this time, through The abnormal warning module sends out an abnormal warning signal to notify the driver. In addition, if the position and range of the abnormal point or abnormal range in the image to be compared is located on the opposite track, the abnormal warning signal issued by the abnormal warning module can notify the driver of this situation, so as to facilitate The abnormal notification lets the driver of the opposite train know.

Preferably, when image difference comparison is performed, if there is an abnormality in the image to be compared after identification and comparison, the abnormality warning module activates a braking unit, and the braking unit reduces the speed of the rail vehicle. When the image recognition module finds that there is an abnormality on the track ahead, such as a vehicle stuck on the track, or a track engineering vehicle forgets to notify when performing operations, the braking unit immediately and automatically performs an appropriate deceleration action according to the current vehicle speed , to ensure that the rail vehicle will not derail or overturn due to sudden braking or hitting a foreign object in front, so as to minimize casualties.

The above-mentioned embodiments are only to illustrate the technical ideas and characteristics of the present invention, and its purpose is to enable those who are familiar with this art to understand the content of the present invention and implement it accordingly, and should not limit the patent scope of the present invention. , all equivalent changes or modifications made in accordance with the spirit of the present invention and the content of the description shall be covered within the patent scope of the present invention.

10:Cloud database module 20: Central monitoring main system 30: On-board subsystem 31: Image capture module 32: Real-time image storage module 33: Data transmission module 34:Reference image storage module 35: Image recognition module 350: ranging unit 36: Abnormal warning module 360: instant notification unit 37: Control module 370: brake unit RV: rail vehicle IC: want to compare images IR: Reference Image

Fig. 1 is a structure diagram of an embodiment of the smart track monitoring system of the present invention; FIG. 2 is a structural diagram of another embodiment of the smart track monitoring system of the present invention; Fig. 3 is a schematic diagram of foreign matter obstructing the track of one embodiment of the smart track monitoring system of the present invention; Fig. 4 is a schematic diagram of image comparison of an embodiment of the smart track monitoring method of the present invention; Fig. 5 is a schematic diagram of an abnormal state of an embodiment of the smart track monitoring method of the present invention; FIG. 6 is a schematic diagram of an abnormal state of another embodiment of the smart track monitoring method of the present invention.

10:Cloud database module

20: Central monitoring main system

30: On-board subsystem

31: Image capture module

32: Real-time image storage module

33: Data transmission module

34:Reference image storage module

35: Image recognition module

36: Image recognition module

37: Image recognition module

RV: rail vehicle

IC: want to compare images

IR: Reference Image

Claims (10)

A smart track monitoring system, comprising: A cloud database module (10), storing the driving record image data of rail vehicles; A central monitoring main system (20), electrically connected to the cloud database module (10), for data signal transmission; and at least one vehicle-mounted subsystem (30), each of which is installed on a plurality of tracks On the vehicle (RV), and electrically connected to the cloud database module (10) and the central monitoring main system (20), respectively, Wherein, the on-board subsystem (30) includes: An image capture module (31), records the running image of the rail vehicle, generates a real-time driving record image, and captures the real-time driving record image to generate an image to be compared (IC); A real-time image storage module (32), which stores the real-time driving record image and the image to be compared (IC); A data transmission module (33), uploads the real-time driving record image to the cloud database module (10) and downloads a reference image (IR) from the cloud database module (10); A reference image storage module (34), storing the reference image (IR); An image identification module (35), identifying the difference between the image to be compared and the reference image; An abnormal warning module (36), when the image recognition module (35) recognizes that there is an abnormal state between the image to be compared and the reference image, the abnormal warning module (36) sends a warning signal; and A control module (37), electrically connected to the image capture module (31), the real-time image storage module (32), the reference image storage module (33), and the data transmission module (34) , the image recognition module (35) and the abnormal warning module (36), and control the operating mechanism between the modules, wherein the image information of the image to be compared and the reference image includes the date and time of recording and the latitude and longitude coordinates . The intelligent track monitoring system according to Claim 1, wherein the image recognition module (35) further includes a distance measuring unit (350), which calculates the distance between the obstacle in front and the rail vehicle. The intelligent track monitoring system as described in claim 1, wherein, the abnormal warning module (36) further includes an instant notification unit (360), which transmits the warning signal to the central monitoring main system (20) or other rail vehicles . The smart track monitoring system as described in claim 1, wherein the control module (37) further includes a braking unit (370), which is electrically connected to the braking system of the rail vehicle (RV) to lower the rail vehicle (RV) Driving speed. A smart track monitoring method, comprising: downloading a reference image from a cloud database module via a data transfer module, and storing the reference image in a reference image storage module; Obtaining a real-time driving record image from an image capturing module, and then capturing an image to be compared from the real-time driving record image, and storing the image to be compared in a real-time image storage module; Using the data transmission module to transmit the real-time driving record image to the cloud database module for storage; Respectively extract a reference latitude and longitude coordinate and a latitude and longitude coordinate to be compared from the reference image and the image to be compared, and use the reference latitude and longitude coordinate and the latitude and longitude coordinate to be compared to perform image overlapping correction, so that the reference image and the latitude and longitude coordinate to be compared The image comparison reference point of the image is within the coordinate error range, and the image difference comparison between the image to be compared and the frame of the reference image is carried out according to the coordinate position of the image; When performing image difference comparison, if there is any abnormality in the image to be compared after identification and comparison, an abnormality warning module installed on the rail vehicle will send out an abnormality warning signal. The smart track monitoring method as described in claim 5, wherein the earlier recorded real-time driving record image is used as the later reference image. The smart track monitoring method as described in claim 5, wherein, when performing image difference comparison, if there is an abnormality in the picture of the image to be compared after identification and comparison, an abnormal range is defined in the image to be compared , according to the current speed of the rail vehicle and the latitude and longitude coordinates, calculate how long it takes to reach the coordinate position of the different range in a unit of time. The smart track monitoring method as described in Claim 7, wherein, after defining an abnormal range in the image to be compared, it is judged whether it is on the track or not according to the position area of the abnormal range in the image to be compared On opposite tracks or across two-way tracks. The smart track monitoring method as described in claim item 5, wherein, when performing image difference comparison, a comparison range is respectively defined on the frames of the image to be compared and the frame of the reference image, and if the image to be compared is identified and compared If there are abnormal points on the screen of the image, but the position of the abnormal point falls outside the comparison range, it is still judged as non-abnormal. The smart track monitoring method as described in claim 5, wherein, when image difference comparison is performed, if there is an abnormality in the screen of the image to be compared after identification and comparison, the abnormality warning module activates a brake unit, and the brake The unit reduces the speed of the rail vehicle.