CN114581364A - GIS defect detection method based on X-ray imaging and Sobel-SCN - Google Patents

GIS defect detection method based on X-ray imaging and Sobel-SCN Download PDF

Info

Publication number
CN114581364A
CN114581364A CN202111537605.0A CN202111537605A CN114581364A CN 114581364 A CN114581364 A CN 114581364A CN 202111537605 A CN202111537605 A CN 202111537605A CN 114581364 A CN114581364 A CN 114581364A
Authority
CN
China
Prior art keywords
image
gis
scn
defect detection
vector set
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
Application number
CN202111537605.0A
Other languages
Chinese (zh)
Inventor
钱锡颖
岳龙
蒋科若
许欣
丁北平
严凌
陆晓波
沈涛
赵鲁臻
王泓学
赖靖胤
李子楠
权超
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ningbo Transmission And Distribution Construction Co ltd
Ningbo Power Supply Co of State Grid Zhejiang Electric Power Co Ltd
Original Assignee
Ningbo Transmission And Distribution Construction Co ltd
Ningbo Power Supply Co of State Grid Zhejiang Electric Power Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Ningbo Transmission And Distribution Construction Co ltd, Ningbo Power Supply Co of State Grid Zhejiang Electric Power Co Ltd filed Critical Ningbo Transmission And Distribution Construction Co ltd
Priority to CN202111537605.0A priority Critical patent/CN114581364A/en
Publication of CN114581364A publication Critical patent/CN114581364A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G06T7/0004—Industrial image inspection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10—Complex mathematical operations
    • G06F17/16—Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/20—Image enhancement or restoration using local operators
    • G06T5/30—Erosion or dilatation, e.g. thinning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/40—Image enhancement or restoration using histogram techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/70—Denoising; Smoothing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T5/00—Image enhancement or restoration
    • G06T5/90—Dynamic range modification of images or parts thereof
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/12—Edge-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/13—Edge detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/155—Segmentation; Edge detection involving morphological operators
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/60—Analysis of geometric attributes
    • G06T7/66—Analysis of geometric attributes of image moments or centre of gravity
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20024—Filtering details
    • G06T2207/20032—Median filtering

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Mathematical Analysis (AREA)
  • Pure & Applied Mathematics (AREA)
  • Mathematical Optimization (AREA)
  • Computational Mathematics (AREA)
  • Software Systems (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Health & Medical Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Molecular Biology (AREA)
  • Artificial Intelligence (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Geometry (AREA)
  • Biomedical Technology (AREA)
  • Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • Quality & Reliability (AREA)
  • Image Analysis (AREA)
  • Analysing Materials By The Use Of Radiation (AREA)

Abstract

The invention discloses a GIS defect detection method based on X-ray imaging and Sobel-SCN, which comprises the following steps: s1, acquiring an X-ray imaging (X-ray imaging) of the GIS device; s2, carrying out rotation, denoising, enhancement and binarization preprocessing on the image; s4, extracting image edge information based on a Sobel algorithm; s5, perfecting image edge contour information based on mathematical morphology; s6, segmenting the image and calibrating the gravity center of the segmented image; s7, constructing a feature vector set; and S8, adopting a Stochastic Configuration Network (SCN) to input the normalized feature vector set, and classifying the feature vector set to realize GIS defect identification. The invention applies the X-ray image processing technology and SCN algorithm to GIS defect detection, replaces inspection personnel to judge whether the GIS has defects, and can improve the quality and efficiency of GIS defect detection.

Description

GIS defect detection method based on X-ray imaging and Sobel-SCN
Technical Field
The invention relates to the technical field of GIS defect detection, in particular to a GIS defect detection method based on X-ray imaging and Sobel-SCN.
Background
Gas Insulated Switch (GIS) equipment is a key device in a substation, and safe and reliable operation of the GIS equipment is particularly important for high-quality power supply.
The nondestructive testing technology based on X-ray imaging is widely applied to the industrial field due to the advantages of intuition, convenience, high testing efficiency and the like. The X-ray imaging technology and the image recognition technology are applied to the GIS defect detection, the GIS nondestructive defect detection can be realized under the condition of not disassembling the equipment, and the quality and the efficiency of the GIS defect detection are improved.
Disclosure of Invention
The technical problem to be solved by the invention is as follows: the GIS defect detection method based on X-ray imaging and Sobel-SCN can replace inspectors to identify whether GIS has defects, improve the quality and efficiency of GIS defect detection, further ensure the safety of the inspectors, reduce economic loss caused by GIS equipment faults, and solve the problems in the prior detection technology.
The invention provides a GIS defect detection method based on X-ray imaging and Sobel-SCN, which comprises the following steps:
s1, acquiring an X-ray image of the GIS equipment;
s2, carrying out rotation, denoising, enhancement and binarization preprocessing on the image;
s3, extracting image edge information based on a Sobel algorithm;
s4, perfecting image edge contour information based on mathematical morphology;
s5, segmenting the image and calibrating the gravity center of the segmented image;
s6, constructing a feature vector set;
and S7, using SCN, using the normalized feature vector set as input, classifying the feature vector set, and realizing GIS defect identification.
Preferably, in step S2, the image is subjected to rotation processing to obtain a position correction image; carrying out median nonlinear filtering denoising processing on the image to obtain a denoised image; carrying out bright-dark contrast enhancement processing on the image to obtain an enhanced image; the binarization processing of the image is to set the pixel point of the X-ray image to 0 or 255.
Preferably, in step S3, the Sobel operator is applied to perform edge detection on the image, so as to obtain image edge information.
Preferably, in step S4, the image is subjected to an open/close operation using the digital morphology principle, so as to obtain perfect image edge contour information.
Preferably, in step S5, the MHFCM algorithm is applied to segment the image to obtain a target area image; and calibrating the gravity center of the segmented image, and providing a basis for constructing a characteristic vector set.
Preferably, in step S6, a feature vector set of the image contour is established for characterizing the edge contour information of the image.
Preferably, the step of feature vector set comprises:
(1) establishing polar coordinates at the center of gravity of the segmented image;
(2) scanning the profile line with the center of gravity as the center;
(3) dividing the contour line into 360 regions according to the circle center to obtain a group of vector point sets of 360 angles, wherein the group of vector point sets are the characteristic vector sets representing the contour line.
Preferably, in step S7, a randomly configured network (SCN) is applied as a classifier, and the following training is performed:
(1) reading in a sample image and calibrating an image label;
(2) processing the image according to the steps from S1 to S6, extracting contour line characteristic information, and taking the normalized characteristic vector set as the input of the SCN model;
(3) training an SCN model;
(4) and performing GIS defect identification.
The beneficial effects of the invention are: compared with the prior art, the invention applies the X-ray image processing technology and the SCN algorithm to the GIS defect detection, and can realize the GIS nondestructive defect detection without disassembling the equipment. In order to improve the detection speed, the image is preprocessed by using rotation, median nonlinear filtering, image intensity and binarization processing; extracting image edge information in a mode of combining a Sobel edge detection operator and a mathematical morphology algorithm; segmenting the image based on an MHFCM algorithm; constructing a characteristic vector set based on the edge contour information of the segmented image; the SCN classifier is adopted to realize GIS defect identification, the accuracy of defect detection is improved, and the economic loss of a power grid company caused by substation accidents is reduced, so that the aim of operating and maintaining the substation is fulfilled.
Drawings
FIG. 1 is a flowchart of a GIS defect detection method based on X-ray imaging and Sobel-SCN.
Fig. 2 is a block diagram of an SCN network.
Detailed Description
In order to make the aforementioned objects, features and advantages of the present invention comprehensible, embodiments accompanied with figures are described in detail below. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present invention. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein, but rather should be construed as broadly as the present invention is capable of modification in various respects, all without departing from the spirit and scope of the present invention.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.
The invention provides a GIS defect detection method based on X-ray imaging and Sobel-SCN, which comprises the following steps:
and S1, acquiring the X-ray image of the GIS device.
In the step, the X-ray machine is arranged at a GIS fault point, so that the angle of the image shot each time is consistent, and a good condition basis is provided for subsequent image processing.
And S2, performing rotation, denoising, enhancing and binaryzation pretreatment on the image.
The step is to rotate the image to obtain a position correction image; carrying out median nonlinear filtering denoising processing on the image to obtain a denoised image; and carrying out bright-dark contrast enhancement processing on the image to obtain an enhanced image. Providing good conditions for subsequent image processing; and setting the pixel point of the image to be 0 or 255, so that the whole image has an obvious black and white effect.
And S3, extracting image edge information based on a Sobel algorithm.
The Sobel algorithm includes two sets of 3 × 3 matrices, which are respectively horizontal and vertical, and performs a planar convolution with the image to obtain horizontal and vertical luminance difference approximations. The images with lateral and longitudinal edge detection are represented by Gx and Gy, respectively, and the formula is as follows:
Figure BDA0003413020350000051
the approximate calculation formula of the horizontal and vertical gradient of each pixel of the image is as follows:
Figure BDA0003413020350000052
the gradient direction calculation formula is as follows:
Figure BDA0003413020350000053
and S4, perfecting the image edge contour line based on mathematical morphology.
The contour line information is perfected by expanding and corroding the contour line based on mathematical morphology. The method comprises the following steps:
(1) expanding the contour line by using a morphological algorithm to seal the contour line;
(2) removing the isolated point set on the non-contour line;
(3) negating the image to obtain each region to be segmented;
(4) performing expansion processing on the contour line again to obtain a more accurate contour line;
and S5, segmenting the image and calibrating the gravity center of the segmented image.
The MHFCM algorithm is adopted to divide the X-ray image into different areas, each area has certain same characteristics, GIS defect detection is facilitated, and the MHFCM algorithm comprises the following steps:
the maximum iteration number is assumed to be iter _ m, and the cost function threshold is assumed to be min _ J.
(1) Calculating a hue frequency histogram of the X-ray image, wherein the calculation formula is as follows:
Figure BDA0003413020350000061
in the formula (4), a represents an image size, and its value is mxn; h(m,n)The image is represented at H(m,n)The hue value of (d); h(m,n)H ∈ {0,1,2, ·, L }, where L is a tone scale.
(2) Calculating the background hue value by the following calculation formula:
Figure BDA0003413020350000062
in the formula (5), p represents a setting window; k (h) e [ hmax-p/2,hmax+p/2]。
(3) Eliminating background color tone area, and processing image with pixel point of Gj,j∈{1,2,...,n}。
(4) Random initialization membership matrix uij∈(0,1),(i∈{1,2,...,c},j∈{1,2,...,n})。
(5) Calculating c clustering centers ciI ∈ {1, 2., c }, the calculation formula is as follows:
Figure BDA0003413020350000071
in the formula (6), GjRepresenting image pixel points; c. CiRepresenting the cluster center of the fuzzy group i; u. ofij∈(0,1)；dij＝||ci-GjL; m ∈ [1, ∞) is a weighted index.
(6) And calculating the value of the value, wherein the calculation formula is as follows:
Figure BDA0003413020350000072
when the value of the cost function is less than min _ J or the number of iterations exceeds iter _ m, the algorithm stops.
(7) Calculating u in the new membership matrix according to equation (6)ij. And returning to the step 5.
The step of calibrating the gravity center of the segmentation image comprises the following steps:
(1) taking the center of the segmented image as an origin, and establishing a rectangular coordinate system;
(2) calculating coordinate values of points on the contour line (such as (x1, y1), (x2, y2), … …, (xm, ym));
(3) and calculating the figure gravity center of the contour line of the segmented image, wherein the calculation formula is as follows:
Figure BDA0003413020350000073
s6, constructing a feature vector set;
and constructing a feature vector set of the image edge contour, wherein the feature vector set is used for characterizing the edge contour information feature of the image. The step of the feature vector set comprises:
(1) establishing polar coordinates at the center of gravity of the segmented image;
(2) scanning the profile line by taking the gravity center as a center;
(3) dividing the contour line into 360 regions according to the circle center, and obtaining a group of vector point sets of 360 angles, wherein the group of vector point sets are the characteristic vector sets representing the contour line. Wherein, the vector point calculation formula is as follows:
Figure BDA0003413020350000081
in the formula (9), α ∈ (0,360)],α∈N+And alpha is a polar angle; n represents a point on the alpha-direction ray having n contour lines;
Figure BDA0003413020350000082
a vector representing the point i on the alpha-direction ray;
Figure BDA0003413020350000083
the vector sum of points sharing n contour lines on the α -direction ray is shown.
And S7, adopting a random configuration network (SCN), taking the normalized feature vector set as input, classifying the feature vector set, and realizing GIS defect identification.
(1) And adopting SCN, taking the normalized feature vector set as input, and detecting the defects of the GIS. The feature vector set normalization calculation formula is as follows:
Figure BDA0003413020350000084
in the formula (9), α ∈ (0,360)],α∈N+And alpha is a polar angle;
Figure BDA0003413020350000085
Figure BDA0003413020350000086
the normalized mode value is expressed by the vector sum of the points having n contour lines on the alpha-direction ray.
(2) As shown in fig. 2, the structure of the SCN network is consistent with the structure of a single-layer feedforward neural network, but the whole neural network structure is built by taking the Boosting idea as a reference.
In fig. 2, the SCN network structure is divided into an input layer, a hidden layer, and an output layer. Wherein, the number of input nodes of the input layer is 361, which comprises 360 input data and a constant 1 bias correction input
Figure BDA0003413020350000091
To input data. The hidden layer is composed of random basis functions generated according to SCN given rules, and the number of nodes of the hidden layer is automatically increased according to a monitoring mechanism. Information propagation among layers is consistent according to an information propagation method of a BP neural network.
(3) And recording the defects in the sample as positive and the defects as negative, and training the SCN model according to the SCN network training strategy.
(4) Performing GIS defect identification
And reading the X-ray image of the GIS equipment except the sample, and performing classified identification. 70 images are adopted for testing, and the accuracy rate reaches 94.6 percent
The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the scope of the patent of the present invention should be defined by the appended claims and equivalents thereof.

Claims (7)

1. A GIS defect detection method based on X-ray imaging and Sobel-SCN comprises the following steps:
s1, acquiring an X-ray image of the GIS equipment;
s2, carrying out rotation, denoising, enhancement and binarization preprocessing on the image;
s3, extracting image edge information based on a Sobel algorithm;
s4, perfecting image edge contour information based on mathematical morphology;
s5, segmenting the image and calibrating the gravity center of the segmented image;
s6, constructing a feature vector set;
and S7, using SCN, using the normalized feature vector set as input, classifying the feature vector set, and realizing GIS defect identification.
2. The GIS defect detection method according to claim 1, wherein in step S2, the image is subjected to rotation processing to obtain a position correction image; carrying out median nonlinear filtering denoising processing on the image to obtain a denoised image; carrying out bright-dark contrast enhancement processing on the image to obtain an enhanced image; the binarization processing of the image is to set the pixel point of the X-ray image to 0 or 255.
3. The GIS defect detection method of claim 1, wherein in step S3, a Sobel operator is applied to perform edge detection on the image to obtain image edge information.
4. The GIS defect detection method of claim 1 wherein, in step S4, the digital morphology principle is applied to perform an open/close operation on the image to obtain complete image edge contour information.
5. The GIS defect detection method of claim 1, wherein in step S5, the image is segmented by using Fuzzy C-means clustering algorithm (MHFCM) segmentation algorithm of windowed tone histogram to obtain the target area image; and calibrating the gravity center of the segmented image, and providing a basis for establishing a characteristic vector set.
6. The GIS defect detection method of claim 1, wherein in step S6, a feature vector set of the image contour is constructed for characterizing edge contour information of the image, and the step of constructing the feature vector set comprises:
(1) establishing polar coordinates at the center of gravity of the segmented image;
(2) scanning the profile line with the center of gravity as the center;
(3) dividing the contour line into 360 regions according to the circle center, and obtaining a group of 360-angle vector point sets, wherein the group of vector point sets are the characteristic vector sets for representing the contour line.
7. The GIS defect detection method of claim 1 wherein, in step S7, a randomly configured network (SCN) is applied as a classifier and the following training is performed:
(1) reading in a sample image and calibrating an image label;
(2) processing the image according to the steps from S1 to S6, extracting contour line characteristic information, and taking the normalized characteristic vector set as the input of the SCN model;
(3) training an SCN model;
(4) and performing GIS defect identification.
CN202111537605.0A 2021-12-15 2021-12-15 GIS defect detection method based on X-ray imaging and Sobel-SCN Pending CN114581364A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111537605.0A CN114581364A (en) 2021-12-15 2021-12-15 GIS defect detection method based on X-ray imaging and Sobel-SCN

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111537605.0A CN114581364A (en) 2021-12-15 2021-12-15 GIS defect detection method based on X-ray imaging and Sobel-SCN

Publications (1)

Publication Number Publication Date
CN114581364A true CN114581364A (en) 2022-06-03

Family

ID=81771230

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111537605.0A Pending CN114581364A (en) 2021-12-15 2021-12-15 GIS defect detection method based on X-ray imaging and Sobel-SCN

Country Status (1)

Country Link
CN (1) CN114581364A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115330802A (en) * 2022-10-17 2022-11-11 山东大学 Carbon fiber composite material gas cylinder X-ray image debonding defect extraction method

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115330802A (en) * 2022-10-17 2022-11-11 山东大学 Carbon fiber composite material gas cylinder X-ray image debonding defect extraction method

Similar Documents

Publication Publication Date Title
CN110543837B (en) Visible light airport airplane detection method based on potential target point
WO2019104767A1 (en) Fabric defect detection method based on deep convolutional neural network and visual saliency
CN106846316B (en) Automatic identification method for typical defect images in GIS
CN110264448B (en) Insulator fault detection method based on machine vision
CN111079747B (en) Railway wagon bogie side frame fracture fault image identification method
CN105067638B (en) Tire fetal membrane face character defect inspection method based on machine vision
CN107808161B (en) Underwater target identification method based on optical vision
CN111091544B (en) Method for detecting breakage fault of side integrated framework of railway wagon bogie
CN111079955A (en) GIS (geographic information System) equipment defect detection method based on X-ray imaging
CN109598681B (en) No-reference quality evaluation method for image after repairing of symmetrical Thangka
CN109523529A (en) A kind of transmission line of electricity defect identification method based on SURF algorithm
CN108830857B (en) Self-adaptive Chinese character copy label image binarization segmentation method
CN117094916B (en) Visual inspection method for municipal bridge support
CN109858570A (en) Image classification method and system, computer equipment and medium
WO2022267270A1 (en) Crack characteristic representation method and system based on multi-fractal spectrum
CN112258490A (en) Low-emissivity coating intelligent damage detection method based on optical and infrared image fusion
CN115830004A (en) Surface defect detection method, device, computer equipment and storage medium
CN110276747B (en) Insulator fault detection and fault rating method based on image analysis
WO2021109011A1 (en) Intelligent capacitor internal defect detection method based on ultrasound image
CN116823725A (en) Aeroengine blade surface defect detection method based on support vector machine
CN114581364A (en) GIS defect detection method based on X-ray imaging and Sobel-SCN
Pratomo et al. Parking detection system using background subtraction and HSV color segmentation
Guo et al. Fault diagnosis of power equipment based on infrared image analysis
CN111402185A (en) Image detection method and device
CN113313678A (en) Automatic sperm morphology analysis method based on multi-scale feature fusion

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