CN112699737A - Genus species identification system and identification method based on biological three-dimensional contour - Google Patents
Genus species identification system and identification method based on biological three-dimensional contour Download PDFInfo
- Publication number
- CN112699737A CN112699737A CN202011454274.XA CN202011454274A CN112699737A CN 112699737 A CN112699737 A CN 112699737A CN 202011454274 A CN202011454274 A CN 202011454274A CN 112699737 A CN112699737 A CN 112699737A
- Authority
- CN
- China
- Prior art keywords
- genus
- imager
- laser scanning
- species
- image
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
- G06V20/41—Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/52—Surveillance or monitoring of activities, e.g. for recognising suspicious objects
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Multimedia (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Microscoopes, Condenser (AREA)
- Investigating, Analyzing Materials By Fluorescence Or Luminescence (AREA)
Abstract
The invention discloses a species identification system based on biological three-dimensional contour, which also comprises: the device comprises a laser scanning confocal micro-imager, an image preprocessor and a species discriminator; the detected organism is placed in a laser scanning confocal micro-imager, the laser scanning confocal micro-imager is connected to an image preprocessor, and the image preprocessor is connected to a species identifier; and the laser scanning confocal microscopic imager is also connected with the species discriminator. The invention also discloses a genus and species identification method based on the biological three-dimensional profile. Compared with the prior art, the genus identification system provided by the application can completely acquire the three-dimensional outline image data of the detected organism, and automatically analyze the data image by using a fine-grained model to obtain the genus classification conclusion of the detected organism, so that the identification efficiency is greatly improved.
Description
Technical Field
The invention belongs to the technical field of biological genus species identification, and particularly relates to a biological genus species identification system.
Background
In the fields of customs import and export detection, natural environment scientific investigation, sample microorganism measurement and the like, the species of organisms to be detected is generally required to be identified. The identification of biological species always troubles experts, firstly, because of the numerous species of organisms, the identification personnel is required to have strong knowledge storage of biological taxonomy, secondly, organisms of different species of the same subject often have many similar biological traits, the biological traits are only slightly different, and the final species identification result is often determined by the slight difference, so that the identification personnel is required to comprehensively master the biological trait knowledge of the tested organisms as much as possible and as comprehensive as possible, and finally, a correct biological species identification conclusion can be obtained.
With the development of science and technology, more and more technical means are involved, so that the automatic biological genus identification becomes possible. In the prior art, methods such as gene sequencing, infrared spectrum identification and the like are commonly used for identifying the species of the detected organism.
Taking the chinese patent application with publication number "CN 105734153A" as an example, the method for species identification of plants by using genes disclosed in the application is disclosed; the molecular identification method for authenticating and distinguishing the authenticity and the species of rhinoceros horn products disclosed in the Chinese patent application with the publication number of CN102732634A and the nucleotide sequence and the method for identifying and distinguishing the species and the variety of Chinese yew disclosed in the Chinese patent application with the publication number of CN104762370B relate to methods for identifying and obtaining the species conclusion of the tested organism by adopting a gene detection method, and the methods are widely used in the field of biological species identification.
Although the method for distinguishing the species by using the gene detection is mature in technology and high in accuracy, the detected organism is inevitably damaged due to the fact that a detection sample of the detected organism needs to be prepared before detection, the detection process is complicated and strict, if the detected organism is a microorganism, a specified strain needs to be cultured in advance, and the reasons result in long time consumption and strict requirements on the specialty in the process of biological identification by using the gene detection and poor popularization.
The method of adopting infrared spectrum to discern also applies to the biological species identification field extensively, for example in Chinese patent application document with patent publication number "CN 101556242B" have disclosed a method for utilizing Fourier's infrared spectrum to discern the microorganism, specifically regard a) to cultivate the contrast microorganism; b) collecting an infrared spectrum of a control microorganism; c) establishing a microorganism identification model in one or more spectra within the interval of 3000-2800cm-1 and 1800-700 cm-1; d) culturing the microorganism to be tested under the same conditions as in step a); e) collecting the infrared spectrum of the microorganism to be detected under the same condition as that in the step b); f) and e) determining the attribution of the microorganisms to be detected by substituting the infrared spectrum obtained in the step e) into a microorganism identification model. Although the method overcomes the problem of destroying the detected organism existing in the gene detection method, the infrared spectrum analysis of different samples has great difficulty due to the different infrared absorption problems, and the identification personnel have abundant spectrum reading experience.
In recent years, as artificial intelligence technology is continuously developed, deep learning related technology is also increasingly applied to the field of biological species identification. The identification method for distinguishing the locust species by the image recognition technology is disclosed in Chinese patent application with the patent application number of '202010105669.2', wherein an image of the locust is shot, and the specific body shape parameters, the morphological characters, the wing shape parameters, the morphological characters and other biological characters of the locust presented in the image are compared with the standard parameters corresponding to the specific locust species in the recognition model, so as to finally obtain the specific conclusion of the locust species displayed in the image. The above patent application document does not note the specific architecture of clear image acquisition, image processing, recognition model and its model training and recognition method, and those skilled in the art cannot know how to implement automatic image recognition from the patent document.
Disclosure of Invention
In order to solve the above problems, an object of the present invention is to provide a genus identification system based on a biological three-dimensional contour, which utilizes confocal microscopic imaging technology to obtain the three-dimensional contour of a detected organism, and sends image data to a genus identifier after imaging, and automatically identifies the species of the detected organism in an image by fine-grained image identification technology, and the system has the advantages of simple structure, high efficiency and reliability in operation, good feasibility, high automation degree, and capability of reducing labor input to the maximum extent.
The invention also aims to provide a biological-based three-dimensional contour identification method, which trains an initial convolutional neural network in a countermeasure mode to obtain a fine-grained model, effectively improves the attention of the model to fine biological characters through training, increases the inter-class distance, reduces the intra-class distance, and can effectively improve the classification precision of image data and obviously improve the accuracy of the classification conclusion of the species of the detected organism when applied to the species identification occasion of the organism.
In order to achieve the purpose, the technical scheme of the invention is as follows:
a genus identification system based on a biological three-dimensional contour, the system further comprising:
the laser scanning confocal micro-imager is used for carrying out laser scanning on the detected organism and carrying out micro-imaging on the detected organism by a confocal imaging technology; an image preprocessor for preprocessing an image generated by the subject organism to generate a challenge sample; a species discriminator with a built-in deep learning analysis model for automatically analyzing the image to obtain species conclusions of organisms contained in the image; the detected organism is placed in a laser scanning confocal micro-imager, the laser scanning confocal micro-imager is connected to an image preprocessor, and the image preprocessor is connected to a species identifier; and the laser scanning confocal microscopic imager is also connected with the species discriminator.
The genus identification system provided by the application can be conveniently applied to biological genus identification occasions with complex biological characteristic combinations for insects, herbaceous plants, fungi, prokaryotes and the like. The laser scanning confocal micro-imager is provided with a laser light source to complete the three-dimensional contour scanning imaging of the detected organism on the basis of a traditional fluorescence microscope, so that the non-invasive observation of the detected organism can be conveniently carried out, the acquired image data can also three-dimensionally, comprehensively, meticulously and clearly contain the biological properties of each dimension of the detected organism, and comprehensive analysis data is provided for subsequent confrontation sample generation and automatic image analysis.
Further, the confocal laser scanning micro-imager comprises: the object stage is used for movably bearing the detected organism; a laser generator for generating laser light; an illumination diaphragm for diffusing the laser beam into a point light source; a bidirectional spectroscope for changing a light propagation path depending on a wavelength; an objective lens for converging incident light rays onto a focal plane; a detection diaphragm for converging the reflected light on the focusing plane; and, a photomultiplier tube for sensing the light signal;
the objective table is provided with a biological placing position to be detected, and the objective table, the objective lens, the two-way spectroscope, the detection diaphragm and the photomultiplier are arranged in sequence; the laser generator and the two-way spectroscope are respectively arranged on two sides of the illumination diaphragm; the light is emitted by the laser generator and forms an incident light path through the illumination diaphragm, the two-way spectroscope and the objective lens; the light rays are emitted by the detected organism to form a reflection light path through the objective lens, the two-way spectroscope, the detection diaphragm and the photomultiplier; during imaging, the illumination diaphragm and the detection diaphragm keep conjugate relative to a focal plane.
After the detected organism is subjected to fluorescence treatment, the detected organism is placed on an object stage, a laser generator is started, the laser generator generates laser beams, the laser beams become point light sources after passing through an illumination diaphragm, the point light sources are refracted by a two-way spectroscope and then irradiate the fluorescence-marked detected organism, a well-defined light spot is formed on a focal plane, and the light spot passes through a detection diaphragm through an objective lens and the two-way spectroscope and then is sensed by a photomultiplier tube.
In the imaging process, the illumination diaphragm and the detection diaphragm keep conjugation relative to a focal plane, namely, a reasonable light path adjusting means is adopted, light spots on the focal plane can be converged on the illumination diaphragm and the detection diaphragm at the same time, fluorescent light spots outside the focal plane can not penetrate through the detection diaphragm and can not be obtained by the induction of a photomultiplier tube due to the difference of focal lengths, therefore, the real-time position of the objective table is changed, the real-time position of the objective table is adjusted, the intersection relation between the focal plane and the detected organism is changed, the detected organism is sliced by the focal plane, the interference of the fluorescent light rays outside the focal plane on the imaging result can be effectively avoided, and the comprehensive, three-dimensional and detailed image data of the detected organism can be obtained.
Furthermore, the laser scanning confocal micro-imager also comprises an incidence filter used for filtering light and converting the point light source into a parallel light source, and the incidence filter is arranged between the illumination diaphragm and the two-way spectroscope.
Furthermore, the confocal laser scanning micro-imager also comprises a reflecting filter used for filtering light and converting the reflected light into parallel light, and the reflecting filter is arranged between the detection diaphragm and the photomultiplier.
The application also discloses a genus identification method based on the biological three-dimensional contour, which comprises the following steps:
s1: modeling: establishing an initial convolutional neural network in a species discriminator;
s2: training a model; inputting training image data in a training image set into an image preprocessor to generate a confrontation sample, and inputting the confrontation sample and original training image data into a species discriminator together to carry out confrontation training on an initial convolutional neural network to obtain a fine-grained model;
s3: imaging; after being processed, the detected living things are placed in a laser scanning confocal micro-imager to obtain image data of the detected living things;
s4: image identification: the image data of the tested living being is input into the genus discriminator to obtain the genus classification prediction conclusion of the tested living being.
Once the training of the initial convolutional neural network is finished, the genus analysis of the detected organism can be automatically carried out in the subsequent image analysis process, and the genus classification prediction conclusion of the detected organism is obtained, so that the efficiency of identifying the genus of the organism is greatly improved, and the labor cost is reduced. In order to obtain higher biological genus species identification accuracy, the identification method provided by the application adopts a countertraining mode to train the original convolutional neural network in advance: the method specifically comprises the following steps:
s21: sequentially placing biological samples with known species conclusions in a laser scanning confocal micro-imager to obtain training image data;
s22: replacing the biological sample, repeating S21 to obtain a plurality of training image data sets, and using the sets as training atlas;
s23: inputting training image data in a training image set into an image preprocessor;
s24: the image preprocessor divides the training image into a plurality of small-size images in a gridding way, randomly scrambles the small-size images, and then splices the small-size images into images with the same size as the original training image as a countermeasure sample;
s25: inputting the confrontation sample and original training image data into a genus discriminator together to carry out confrontation training on the initial convolutional neural network;
s26; and repeating 23-25 until the classification accuracy of the convolutional neural network meets the requirement, and keeping the network parameter output to be a fine-grained model.
The method is characterized in that a training image is segmented in a gridding mode and randomly disturbed, the influence of noise caused by image damage on a model is reduced by using a countermeasure method, namely, on the basis of the original classification, classification categories are doubled, a convolutional neural network is required to identify the category corresponding to the original training image and the category corresponding to a countermeasure sample, so that the original convolutional neural network model focuses more on fine-grained characteristics, and compared with a traditional model, the finally obtained fine-grained model focuses more on the biological characteristics of the detected organism and has higher species classification prediction accuracy.
The invention has the advantages that: compared with the prior art, the genus identification system provided by the application can completely acquire the three-dimensional outline image data of the detected living beings, and automatically analyzes the data image by using the fine-grained model to obtain the genus classification conclusion of the detected living beings, so that the identification efficiency is greatly improved.
Drawings
Fig. 1 is a system architecture diagram of a species identification system based on a biological three-dimensional profile provided in an embodiment.
Fig. 2 is a schematic structural diagram of a laser scanning confocal micro-imager in a species identification system based on a biological three-dimensional profile provided in an embodiment.
Fig. 3 is an optical path diagram of a confocal laser scanning microscopy imager in a biological-based species identification system during imaging according to an embodiment, wherein solid lines represent incident light rays and dotted lines represent reflected light rays.
FIG. 4 is a flowchart of a genus identification method based on a biological three-dimensional contour according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the detailed description and specific examples, while indicating the invention, are intended for purposes of illustration only and are not intended to limit the scope of the invention.
In order to achieve the purpose, the technical scheme of the invention is as follows:
please refer to fig. 1-3.
In this embodiment, a genus identification system based on a three-dimensional biological contour is provided, which further comprises:
a laser scanning confocal micro-imager 1 for performing laser scanning on the detected organism and performing micro-imaging on the detected organism by a confocal imaging technology; an image preprocessor 2 for preprocessing an image generated by the subject organism to generate a challenge sample; a genus discriminator 3 with a built-in deep learning analysis model for automatically analyzing the image to obtain a genus conclusion of the living beings contained in the image; the detected organism is arranged in a laser scanning confocal micro-imager 1, the laser scanning confocal micro-imager 1 is connected to an image preprocessor 2, and the image preprocessor 2 is connected to a species identifier 3; and the laser scanning confocal micro-imager 1 is also connected with a species discriminator 3.
In the present embodiment, the laser scanning confocal micro-imager 1 includes: a stage 11 for movably carrying a subject organism; a laser generator 12 for generating laser light; an illumination diaphragm 13 for diffusing the laser beam into a point light source; a two-way beam splitter 14 for changing the propagation path of light depending on the wavelength; an objective lens 15 for converging incident light rays onto a focal plane; a detection diaphragm 16 for converging the reflected light on the focal plane; and, a photomultiplier 17 for sensing the light signal;
the biological subject is disposed on the objective table 11 after being processed, the objective table 11, the objective lens 15, the two-way spectroscope 14, the detection diaphragm 16 and the photomultiplier 17 are sequentially disposed; the laser generator 12 and the two-way beam splitter 14 are respectively arranged at two sides of the illumination diaphragm 13; light is emitted by a laser generator 12 and forms an incident light path through an illumination diaphragm 13, a two-way beam splitter 14 and an objective lens 15; the light is emitted by the detected organism and forms a reflection light path through an objective lens 15, a two-way spectroscope 14, a detection diaphragm 16 and a photomultiplier 17; during imaging, the illumination diaphragm 13 and the detection diaphragm 16 remain conjugated with respect to the focal plane.
In the present embodiment, the confocal laser scanning micro-imager 1 further includes an entrance filter 18 for filtering and converting the point light source into a parallel light source, and the entrance filter 18 is disposed between the illumination diaphragm 13 and the two-way beam splitter 14.
In the present embodiment, the confocal laser scanning micro-imager 1 further includes a reflective filter 19 for filtering and converting the reflected light into parallel light, and the reflective filter 19 is disposed between the detection diaphragm 16 and the photomultiplier 17.
In this embodiment, a genus identification method based on a biological three-dimensional contour is also disclosed, which comprises:
s1: modeling: establishing an initial convolutional neural network in the genus discriminator 3;
s2: training a model; inputting training image data in a training image set into an image preprocessor 2 to generate a confrontation sample, and inputting the confrontation sample and original training image data into a species discriminator 3 together to carry out confrontation training on an initial convolutional neural network to obtain a fine-grained model;
s3: imaging; after being processed, the detected living things are placed in a laser scanning confocal micro-imager 1 to obtain the image data of the detected living things;
s4: image identification: the image data of the test subject is inputted to the genus discriminator 3 to obtain the genus classification prediction conclusion of the test subject.
Further, in this embodiment, S2 is specifically:
s21: sequentially placing biological samples with known species conclusions in a laser scanning confocal micro-imager 1 to obtain training image data;
s22: replacing the biological sample, repeating S21 to obtain a plurality of training image data sets, and using the sets as training atlas;
s23: inputting training image data in a training image set into an image preprocessor 2;
s24: the image preprocessor 2 divides the training image into a plurality of small-size images in a gridding way, randomly scrambles the small-size images, and then splices the small-size images into images with the same size as the original training image as a countermeasure sample;
s25: inputting the confrontation sample and the original training image data into a genus discriminator 3 together to carry out confrontation training on the initial convolutional neural network;
s26; and repeating 23-25 until the classification accuracy of the convolutional neural network meets the requirement, and keeping the network parameter output to be a fine-grained model.
The present invention is not limited to the above preferred embodiments, and any modifications, equivalent substitutions and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (6)
1. A genus identification system based on a biological three-dimensional profile, the system further comprising:
the laser scanning confocal micro-imager is used for carrying out laser scanning on the detected organism and carrying out micro-imaging on the detected organism by a confocal imaging technology;
an image preprocessor for preprocessing an image generated by the subject organism to generate a challenge sample;
a species discriminator with a built-in deep learning analysis model for automatically analyzing the image to obtain species conclusions of organisms contained in the image;
a detected organism placing position is arranged in the laser scanning confocal micro-imager, the laser scanning confocal micro-imager is connected to the image preprocessor, and the image preprocessor is connected to the species identifier;
and the laser scanning confocal microscopic imager is also connected with the genus identifier.
2. The biometric-based genus identification system of claim 1, wherein said confocal laser scanning microscopy imager comprises:
the object stage is used for movably bearing the detected organism;
a laser generator for generating laser light;
an illumination diaphragm for diffusing the laser beam into a point light source;
a bidirectional spectroscope for changing a light propagation path depending on a wavelength;
an objective lens for converging incident light rays onto a focal plane;
a detection diaphragm for converging the reflected light on the focusing plane;
and, a photomultiplier tube for sensing the light signal;
the objective table, the objective lens, the two-way spectroscope, the detection diaphragm and the photomultiplier are arranged in sequence; the laser generator and the two-way spectroscope are respectively arranged on two sides of the illumination diaphragm;
the light is emitted by the laser generator and forms an incident light path through the illumination diaphragm, the two-way beam splitter and the objective lens;
the light rays are emitted by the detected organism to form a reflection light path through the objective lens, the two-way spectroscope, the detection diaphragm and the photomultiplier;
during imaging, the illumination diaphragm and the detection diaphragm keep conjugate relative to a focal plane.
3. The biometric-based genus identification system of claim 2, wherein said confocal laser scanning microscopy imager further comprises an entrance filter for filtering and converting point light sources into collimated light sources, said entrance filter being disposed between said illumination aperture and said two-way beam splitter.
4. The biobased stereotopographic genus species identification system of claim 3, wherein said confocal laser scanning micro-imager further comprises a reflecting filter for filtering and converting reflected light into parallel light, said reflecting filter being disposed between said detection aperture and said photomultiplier tube.
5. A genus identification method based on biological three-dimensional contour is characterized by comprising the following steps:
s1: modeling: establishing an initial convolutional neural network in the genus discriminator;
s2: training a model; inputting training image data in a training image set into the image preprocessor to generate a confrontation sample, and inputting the confrontation sample and original training image data into the genus discriminator together to carry out confrontation training on the initial convolutional neural network to obtain a fine-grained model;
s3: imaging; after being processed, the detected living things are placed in the laser scanning confocal micro-imager to obtain the image data of the detected living things;
s4: image identification: inputting the image data of the detected organism into the genus discriminator to obtain the genus classification prediction conclusion of the detected organism.
6. The genus species identification method based on the biological three-dimensional contour according to claim 5, wherein said S2 is specifically:
s21: sequentially placing biological samples with known species conclusions in the laser scanning confocal micro-imager to obtain training image data;
s22: replacing the biological sample, repeating S21 to obtain a plurality of training image data sets, and using the sets as training atlas;
s23: inputting training image data in a training image set into the image preprocessor;
s24: the image preprocessor divides the training image into a plurality of small-size images in a gridding way, randomly scrambles the small-size images, and then splices the small-size images into images with the same size as the original training image as a countermeasure sample;
s25: inputting the confrontation sample and original training image data into the genus discriminator together to carry out confrontation training on the initial convolutional neural network;
s26; and repeating 23-25 until the classification accuracy of the convolutional neural network meets the requirement, and keeping the network parameter output to be a fine-grained model.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011454274.XA CN112699737A (en) | 2020-12-10 | 2020-12-10 | Genus species identification system and identification method based on biological three-dimensional contour |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011454274.XA CN112699737A (en) | 2020-12-10 | 2020-12-10 | Genus species identification system and identification method based on biological three-dimensional contour |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112699737A true CN112699737A (en) | 2021-04-23 |
Family
ID=75508631
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011454274.XA Pending CN112699737A (en) | 2020-12-10 | 2020-12-10 | Genus species identification system and identification method based on biological three-dimensional contour |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112699737A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120526423A (en) * | 2025-05-21 | 2025-08-22 | 榕城海关综合技术服务中心 | An efficient insect identification system based on the integration of microscopic images and artificial intelligence |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018117273A1 (en) * | 2016-12-22 | 2018-06-28 | 国立大学法人筑波大学 | Data creation method and data use method |
| US20190011882A1 (en) * | 2017-07-05 | 2019-01-10 | Accelerate Diagnostics, Inc. | Lens-free holographic optical system for high sensitivity label-free microbial growth detection and quantification for screening, identification, and susceptibility testing |
| US20190333199A1 (en) * | 2018-04-26 | 2019-10-31 | The Regents Of The University Of California | Systems and methods for deep learning microscopy |
| CN111444927A (en) * | 2020-03-26 | 2020-07-24 | 广州市炜城智能科技有限公司 | Method for identifying animal and plant pests by using laser technology in multidimensional manner |
| CN111860327A (en) * | 2020-07-21 | 2020-10-30 | 广州道源信息科技有限公司 | Image detection and analysis method based on visual computation video transmission |
| CN111914928A (en) * | 2020-07-30 | 2020-11-10 | 南京大学 | A method for adversarial example defense for image classifiers |
| CN112052761A (en) * | 2020-08-27 | 2020-12-08 | 腾讯科技（深圳）有限公司 | Method and device for generating confrontation face image |
-
2020
- 2020-12-10 CN CN202011454274.XA patent/CN112699737A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2018117273A1 (en) * | 2016-12-22 | 2018-06-28 | 国立大学法人筑波大学 | Data creation method and data use method |
| US20190011882A1 (en) * | 2017-07-05 | 2019-01-10 | Accelerate Diagnostics, Inc. | Lens-free holographic optical system for high sensitivity label-free microbial growth detection and quantification for screening, identification, and susceptibility testing |
| US20190333199A1 (en) * | 2018-04-26 | 2019-10-31 | The Regents Of The University Of California | Systems and methods for deep learning microscopy |
| CN111444927A (en) * | 2020-03-26 | 2020-07-24 | 广州市炜城智能科技有限公司 | Method for identifying animal and plant pests by using laser technology in multidimensional manner |
| CN111860327A (en) * | 2020-07-21 | 2020-10-30 | 广州道源信息科技有限公司 | Image detection and analysis method based on visual computation video transmission |
| CN111914928A (en) * | 2020-07-30 | 2020-11-10 | 南京大学 | A method for adversarial example defense for image classifiers |
| CN112052761A (en) * | 2020-08-27 | 2020-12-08 | 腾讯科技（深圳）有限公司 | Method and device for generating confrontation face image |
Non-Patent Citations (3)
| Title |
|---|
| 刘恒;吴德鑫;徐剑;: "基于生成式对抗网络的通用性对抗扰动生成方法", 信息网络安全, no. 05 * |
| 刘西蒙;谢乐辉;王耀鹏;李旭如;: "深度学习中的对抗攻击与防御", 网络与信息安全学报, no. 05 * |
| 潘文雯等: "对抗样本生成技术综述", 《软件学报》, vol. 31, no. 01, pages 1 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120526423A (en) * | 2025-05-21 | 2025-08-22 | 榕城海关综合技术服务中心 | An efficient insect identification system based on the integration of microscopic images and artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11542461B2 (en) | Analysis device | |
| JP7336654B2 (en) | Processing system, data processing device, display system and microscope system | |
| CN109528161B (en) | High-speed high-resolution scanning microscopic imaging system and method based on machine learning | |
| US11598712B2 (en) | System and method for cell evaluation, and cell evaluation program | |
| US9013692B2 (en) | Flow cytometer apparatus for three dimensional difraction imaging and related methods | |
| CN110389119B (en) | Quick self-adaptive optical scanning microscopic imaging system and method based on machine learning | |
| CN103154708A (en) | Method for measuring diffusion characteristic value of particle by detecting single light-emitting particle | |
| Balmages et al. | Laser speckle imaging for early detection of microbial colony forming units | |
| CN110596059A (en) | Optical super-resolution microscopic imaging system | |
| Kayan et al. | Intensity and phase stacked analysis of a Φ-OTDR system using deep transfer learning and recurrent neural networks | |
| Polat et al. | High-precision laser focus positioning of rough surfaces by deep learning | |
| CN112699737A (en) | Genus species identification system and identification method based on biological three-dimensional contour | |
| CN105842224B (en) | Blood discrimination method based on infrared Raman Ultraluminescence super continuous spectrums | |
| Hu et al. | Enhancing vibration detection in Φ-OTDR through image coding and deep learning-driven feature recognition | |
| Feizpour et al. | Bacterial identification in SERS-integrated microfluidics using CNN-driven 2D classification of 1D spectra | |
| CN111239993B (en) | Super-resolution total internal reflection microscopic imaging device and method based on polar scattering | |
| CN119985445A (en) | A method for detecting contamination on the surface of silkworm cocoons | |
| Yuliastuti et al. | Compact computer vision system for tropical wood species recognition based on pores and concentric curve | |
| Ong et al. | Phantom validation of Monte Carlo modeling for noncontact depth sensitive fluorescence measurements in an epithelial tissue model | |
| Yetiş et al. | Myelin segmentation in fluorescence microscopy images | |
| Zhang et al. | Optical time-series signals classification based on data augmentation for small sample | |
| CN115128052A (en) | Water pollution tracing method for three-dimensional fluorescence signal PCA-neural network recognition | |
| Yoon et al. | Label-free identification of non-activated lymphocytes using three-dimensional refractive index tomography and machine learning | |
| CN105842177B (en) | Blood discrimination method based on the super continuous unrestrained comprehensive spectrum of Ultraluminescence | |
| Zbezhkhovska | DEEPFAKE AUDIO DETECTION USING YOLOV8 WITH MEL-SPECTROGRAM ANALYSIS: A CROSS-DATASET EVALUATION |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210423 |