CN116823756A - Pile leg weld defect detection method - Google Patents
Pile leg weld defect detection method Download PDFInfo
- Publication number
- CN116823756A CN116823756A CN202310745818.5A CN202310745818A CN116823756A CN 116823756 A CN116823756 A CN 116823756A CN 202310745818 A CN202310745818 A CN 202310745818A CN 116823756 A CN116823756 A CN 116823756A
- Authority
- CN
- China
- Prior art keywords
- image
- defect
- area
- detection method
- weld
- 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
- 230000007547 defect Effects 0.000 title claims abstract description 115
- 238000001514 detection method Methods 0.000 title claims abstract description 63
- 238000000034 method Methods 0.000 claims abstract description 33
- 230000009466 transformation Effects 0.000 claims abstract description 12
- 238000013527 convolutional neural network Methods 0.000 claims abstract description 8
- 238000003708 edge detection Methods 0.000 claims abstract description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 6
- 238000006243 chemical reaction Methods 0.000 claims description 5
- 230000005484 gravity Effects 0.000 claims description 5
- 238000012549 training Methods 0.000 claims description 5
- 230000008859 change Effects 0.000 claims description 4
- 238000003066 decision tree Methods 0.000 claims description 4
- 238000001914 filtration Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 230000001131 transforming effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 239000002893 slag Substances 0.000 description 3
- 238000009826 distribution Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 230000001186 cumulative effect Effects 0.000 description 1
- 238000013501 data transformation Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
- 238000003466 welding Methods 0.000 description 1
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to a pile leg weld defect detection method, which utilizes a wavelet transformation mode maximum multi-scale edge detection method to acquire the image edge of a maximum connected domain, thereby realizing the acquisition of a potential area with defects. And judging whether the potential area contains defects or not through a weld joint area division model, and judging specific defect areas and defect types of the defects through a convolutional neural network model. Compared with the existing weld defect detection method, the method has the advantages that the non-weld region is removed by reserving the connected region with the largest area, and the detection accuracy is improved. Meanwhile, before the defect area and the defect category are identified, whether the potential area contains the defect is judged in advance through a weld joint area division model, so that the accuracy of defect detection is further improved, missing detection and false detection of the defect are reduced, and the effectiveness of weld joint defect detection is greatly improved.
Description
Technical Field
The invention relates to a weld defect detection method, in particular to a pile leg weld defect detection method.
Background
Along with the rapid development of the technology of identifying weld defects by targets, behavior simulation and big data technology, a method for detecting pile leg weld defects by using a deep learning method gradually becomes a mainstream, and the method has wide application in the traditional industrial vision field.
Aiming at the problem of pile leg weld defect detection, researchers in countries around the world develop related researches, wherein the method for detecting the weld defect by using the rapid detection and segmentation combined by various threshold modes, which is proposed by Abdehak M, specifically adopts an X-ray image and combines with a threshold method to extract the weld defect characteristics, and the algorithm has higher detection accuracy for a relatively clear weld defect area. However, when the background of the image is complex and changeable, for example, the brightness imbalance phenomenon exists in the image, the detection effect will be reduced linearly, so Zhichao Liao et al propose a method for detecting a fixed background, the main idea is to fit the background area existing in detection, and finally complete the extraction of the weld defect characteristics of the detection area by removing the background area.
Although the above weld defect detection method has good detection effect on weld defect detection, there are a number of disadvantages, for example, although the weld defect detection area can be roughly positioned, there is a certain position difference with the true defect area, and the condition of missing detection and false detection can occur in the detection. Meanwhile, in the algorithm design of the method, the same points of the defect image and the defect-free image are not considered, the defect image and the defect-free image exist uniformly in actual production, and the defect image detected in the weld-free image possibly appears in detection, so that the detection accuracy is greatly influenced, the universality of the method is weaker, and a better weld defect detection algorithm is required to be provided under different actual production scenes to improve the detection effect.
The method for identifying the semitrailer axle defect based on the computer vision comprises the steps of determining an edge rule degree value of each suspicious connected domain by acquiring each suspicious connected domain in a weld joint area image of the axle surface, and further determining each target connected domain in each suspicious connected domain; according to the gray values of the pixel points in each target connected domain, determining the internal gray change degree value and gray curve evaluation value corresponding to each target connected domain, and further determining the credibility degree of slag inclusion defects corresponding to each target connected domain, so as to finally determine the slag inclusion defect areas in each target connected domain. The defect identification method can realize detection of the welding seam defect, but is mainly aimed at slag inclusion defects and air hole defects, and has limited application range.
Disclosure of Invention
In order to solve the technical problems, the invention aims to provide the pile leg weld defect detection method which is not easy to misdetect and has high detection accuracy.
The pile leg weld defect detection method provided by the invention comprises the following steps of:
step S1, a step S1; acquiring potential areas in the image which may have defects;
s2, a step of S2; judging whether the potential area obtained in the step S1 has defects or not by using the trained weld joint area dividing model, and outputting the potential area with the defects;
s3, a step of S3; judging a specific defect area and a defect category in the potential area output in the step S2 by using the trained convolutional neural network model;
the step of acquiring the potential area with the possible defect in the step S1 is as follows;
step S11; extracting all connected domains in the image;
step S12; reserving the connected domain with the largest area, and setting the gray value in the rest connected domains to be 0;
step S13; for the connected domain with the largest area in the step S12, an image edge is obtained by using a wavelet transformation mode maximum multi-scale edge detection method, and a region containing the image edge is selected by a frame, wherein the region is a potential region with a possible defect.
The pile leg weld defect detection method has the advantages that the image edge of the largest connected domain is obtained by using a wavelet transformation mode maximum multi-scale edge detection method, and further the acquisition of a potential area with defects is realized. And judging whether the potential area contains defects or not through a weld joint area division model, and judging specific defect areas and defect types of the defects through a convolutional neural network model.
As the weld joint area is generally larger, compared with the existing weld joint defect detection method, the pile leg weld joint defect detection method has the advantages that the communicating region with the largest area is reserved, so that the non-weld joint area is eliminated, the defect detection in the non-weld joint area is avoided, and the detection accuracy is improved.
Meanwhile, before the defect area and the defect category are identified, whether the potential area contains the defects is judged in advance through the weld joint area division model, so that the defect detection efficiency is improved, the defect detection accuracy is further improved, the missing detection and the false detection of the defects are reduced, and the effectiveness of weld joint defect detection is greatly improved.
Furthermore, the pile leg weld defect detection method of the invention carries out pretreatment on the image before the step S1, and comprises the following steps of;
the step of denoising the image: denoising the image by using nonlinear filtering;
a step of performing gradation change on the image: converting the three-channel image into a single-channel image;
a step of transforming the image data: correcting the image using gamma conversion;
and equalizing the image by using the gray level histogram.
The preprocessing of the image facilitates the acquisition of the potential area by the subsequent program, and improves the accuracy of the acquisition of the potential area.
Further, according to the pile leg weld defect detection method, in the step S2, a weld joint region division model is obtained through the following steps;
step S21; data arrangement is carried out on the extracted characteristic data, and a true label is added;
step S22; and (3) training the weld joint region division model by utilizing the characteristic data extracted in the step S21 and through a decision tree classifier.
Furthermore, the pile leg weld defect detection method disclosed by the invention comprises an image normalization step before the step S3;
the image is larger than or equal to 416 x 416 pixels, and scaling is adopted in equal proportion; pixels smaller than 416 x 416 are scaled in unequal proportions by using a team-clamping algorithm;
when the image is subjected to anisometric transformation, the image size is firstly judged, if the image exceeds the normalized size, the image is subjected to the anisometric scaling, and after the specified size requirement is met, the anisometric transformation is continued, so that the image finally meets the size requirement.
Further, the pile leg weld defect detection method provided by the invention further comprises the following steps of;
s14: dividing the potential area into a plurality of columns;
s15: for each column of the potential area, a square weighted gray level gravity center method is utilized to obtain a brightness center point in the column area, and a calculation formula is as follows:
wherein (v) c ,ν c ) Is the coordinates of the luminance center point, (u, v) i ) The pixel point coordinate is a pixel point coordinate in a column, N is the total number of the pixel points in the column, and I is a pixel point gray value;
s16: and (3) fitting the brightness center points in each column region acquired in the step (S15) into continuous sub-pixel level center lines by using continuous edge fitting operation, and calculating the area and the perimeter of the surrounding region of the sub-pixel level center lines.
The center point is obtained by adopting a square weighted gray level gravity center method, and then the sub-pixel level center line and the surrounding area and perimeter thereof are obtained through continuous edge fitting operation, and the weld joint area division model in the step S2 can predict the area and perimeter characteristics to judge whether the weld joint area division model is a defect or not.
The foregoing description is merely an overview of the embodiments of the present invention, and is intended to provide a more clear understanding of the technical means of the present invention, as embodied in the present invention, by way of example only.
Drawings
FIG. 1 is a flow chart of a leg weld defect detection method.
Detailed Description
The following describes in further detail the embodiments of the present invention with reference to the drawings and examples. The following examples are illustrative of the invention and are not intended to limit the scope of the invention.
Referring to fig. 1, the method for detecting the pile leg weld defect in the embodiment includes the following steps:
step S1, a step S1; acquiring potential areas in the image which may have defects;
s2, a step of S2; judging whether the potential area obtained in the step S1 has defects or not by using the trained weld joint area dividing model, and outputting the potential area with the defects;
s3, a step of S3; judging a specific defect area and a defect category in the potential area output in the step S2 by using the trained convolutional neural network model;
the step of acquiring the potential area with the possible defect in the step S1 is as follows;
step S11; extracting all connected domains in the image;
step S12; reserving the connected domain with the largest area, and setting the gray value in the rest connected domains to be 0;
step S13; for the connected domain with the largest area in the step S12, an image edge is obtained by using a wavelet transformation mode maximum multi-scale edge detection method, and a region containing the image edge is selected by a frame, wherein the region is a potential region with a possible defect.
According to the pile leg weld defect detection method, the image edge of the largest connected domain is obtained by using the wavelet transformation mode maximum multi-scale edge detection method, so that the acquisition of a potential area with defects is realized. And judging whether the potential area contains defects or not through a weld joint area division model, and judging specific defect areas and defect types of the defects through a convolutional neural network model.
As the weld joint area is generally larger, compared with the existing weld joint defect detection method, the pile leg weld joint defect detection method has the advantages that the communicating region with the largest area is reserved, so that the non-weld joint area is eliminated, the defect detection in the non-weld joint area is avoided, and the detection accuracy is improved.
Meanwhile, before the defect area and the defect category are identified, whether the potential area contains the defects is judged in advance through the weld joint area division model, so that the defect detection efficiency is improved, the defect detection accuracy is further improved, the missing detection and the false detection of the defects are reduced, and the effectiveness of weld joint defect detection is greatly improved.
The area of the connected domain is the number of pixels with gray values different from 0 in the connected domain.
The wavelet transformation mode maximum multi-scale edge detection method is an existing image edge detection method, and can accurately and rapidly extract image edges.
The weld joint region division model is used for judging whether a potential region has defects or not, and the characteristic information is obtained through characteristic extraction, and then model training is carried out by using a classifier to realize classification, so that the division of correct weld joint defect regions is achieved;
the extracted characteristic data are tidied and truly labeled;
the characteristic data is utilized to predict and classify the data to be detected after model training is carried out through a decision tree classifier;
the convolutional neural network model is used for acquiring a specific defect area and defect category, and is a YOLOv3 network model, and the specific structure and principle of the convolutional neural network model can refer to related documents and are not repeated.
Preferably, the method for detecting the pile leg weld defect of the embodiment pre-processes an image before step S1, including the following steps;
the step of denoising the image: denoising the image by using nonlinear filtering;
a step of performing gradation change on the image: converting the three-channel image into a single-channel image;
a step of transforming the image data: correcting the image using gamma conversion;
and equalizing the image by using the gray level histogram.
The preprocessing of the image facilitates the acquisition of the potential area by the subsequent program, improves the accuracy of the acquisition of the potential area,
the step of carrying out data transformation on the image enhances the contrast ratio of the target and the background, specifically, the gamma transformation corrects the gray value, after the value which is too low or too high is calculated, the gray value needs to be assigned and calculated again, the product operation is carried out on the pixel value and the parameter of each position on the original image, and the enhanced visual effect of the image can be better.
The method for denoising the image by nonlinear filtering, converting the three-channel image into a single-channel image and correcting the image by gamma conversion is the existing method, and the specific principle and the step process are not repeated.
The step of equalizing the image using the gray histogram is as follows:
a1: determining the gray level of the image so as to judge whether gray conversion is needed;
a2: calculating original histogram distribution probability;
a3: calculating a histogram probability cumulative value;
a4: calculating a formula of pixel mapping;
a5: performing formula mapping on all pixels in the image, wherein all pixels after mapping form a new image;
the step of equalizing the image by using the gray histogram improves the visual effect of the image by expanding the distribution range of the gray of the image.
As an preference, in the pile leg weld defect detection method of the present embodiment, the weld region division model in step S2 is obtained by the following steps;
step S21; data arrangement is carried out on the extracted characteristic data, and a true label is added;
step S22; and (3) training the weld joint region division model by utilizing the characteristic data extracted in the step S21 and through a decision tree classifier.
Preferably, the method for detecting the pile leg weld defect in the embodiment includes an image normalization step before step S3;
the image is larger than or equal to 416 x 416 pixels, and scaling is adopted in equal proportion; pixels smaller than 416 x 416 are scaled in unequal proportions by using a team-clamping algorithm;
when the image is subjected to anisometric transformation, the image size is firstly judged, if the image exceeds the normalized size, the image is subjected to the anisometric scaling, and after the specified size requirement is met, the anisometric transformation is continued, so that the image finally meets the size requirement.
Preferably, the method for detecting the weld defect of the pile leg of the embodiment further comprises the following steps of;
s14: dividing the potential area into a plurality of columns;
s15: for each column of the potential area, a square weighted gray level gravity center method is utilized to obtain a brightness center point in the column area, and a calculation formula is as follows:
wherein (v) c ,ν c ) Is the coordinates of the luminance center point, (u, v) i ) Is thatA certain pixel point coordinate in a column, N is the total number of the pixel points in the column, and I is the gray value of the pixel points;
s16: and (3) fitting the brightness center points in each column region acquired in the step (S15) into continuous sub-pixel level center lines by using continuous edge fitting operation, and calculating the area and the perimeter of the surrounding region of the sub-pixel level center lines.
The center point is obtained by adopting a square weighted gray level gravity center method, the sub-pixel level center line and the surrounding area and circumference thereof are obtained by continuous edge fitting operation, and the weld joint area division model in the step S2 can utilize the area and circumference characteristics to predict the area and circumference characteristics so as to more accurately judge whether the area and circumference characteristics are defects.
The above is only a preferred embodiment of the present invention for assisting a person skilled in the art to implement the corresponding technical solution, and is not intended to limit the scope of the present invention, which is defined by the appended claims. It should be noted that, on the basis of the technical solution of the present invention, several improvements and modifications equivalent thereto can be made by those skilled in the art, and these improvements and modifications should also be regarded as the protection scope of the present invention. Meanwhile, it should be understood that, although the present disclosure describes the above embodiments, not every embodiment contains only one independent technical solution, and the description is merely for clarity, and those skilled in the art should consider the disclosure as a whole, and the technical solutions of the embodiments may be combined appropriately to form other embodiments that can be understood by those skilled in the art.
Claims (5)
1. The pile leg weld defect detection method is characterized by comprising the following steps of:
step S1, a step S1; acquiring potential areas in the image which may have defects;
s2, a step of S2; judging whether the potential area obtained in the step S1 has defects or not by using the trained weld joint area dividing model, and outputting the potential area with the defects;
s3, a step of S3; judging a specific defect area and a defect category in the potential area output in the step S2 by using the trained convolutional neural network model;
the step of acquiring the potential area with the possible defect in the step S1 is as follows;
step S11; extracting all connected domains in the image;
step S12; reserving the connected domain with the largest area, and setting the gray value in the rest connected domains to be 0;
step S13; for the connected domain with the largest area in the step S12, an image edge is obtained by using a wavelet transformation mode maximum multi-scale edge detection method, and a region containing the image edge is selected by a frame, wherein the region is a potential region with a possible defect.
2. The leg weld defect detection method according to claim 1, wherein: preprocessing the image before the step S1, wherein the preprocessing comprises the following steps of;
the step of denoising the image: denoising the image by using nonlinear filtering;
a step of performing gradation change on the image: converting the three-channel image into a single-channel image;
a step of transforming the image data: correcting the image using gamma conversion;
and equalizing the image by using the gray level histogram.
3. The leg weld defect detection method according to claim 1, wherein: the weld joint region division model in the step S2 is obtained through the following steps;
step S21; data arrangement is carried out on the extracted characteristic data, and a true label is added;
step S22; and (3) training the weld joint region division model by utilizing the characteristic data extracted in the step S21 and through a decision tree classifier.
4. The leg weld defect detection method according to claim 1, wherein: the step S3 is preceded by an image normalization step;
the image is larger than or equal to 416 x 416 pixels, and scaling is adopted in equal proportion; pixels less than 416 x 416 are scaled non-uniformly using the sea-Carving algorithm.
5. The leg weld defect detection method according to claim 1, wherein: the method also comprises the following steps of;
s14: dividing the potential area into a plurality of columns;
s15: for each column of the potential area, a square weighted gray level gravity center method is utilized to obtain a brightness center point in the column area, and a calculation formula is as follows:
wherein (v) c ,ν c ) Is the coordinates of the luminance center point, (u, v) i ) The pixel point coordinate is a pixel point coordinate in a column, N is the total number of the pixel points in the column, and I is a pixel point gray value;
s16: and (3) fitting the brightness center points in each column region acquired in the step (S15) into continuous sub-pixel level center lines by using continuous edge fitting operation, and calculating the area and the perimeter of the surrounding region of the sub-pixel level center lines.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310745818.5A CN116823756A (en) | 2023-06-25 | 2023-06-25 | Pile leg weld defect detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310745818.5A CN116823756A (en) | 2023-06-25 | 2023-06-25 | Pile leg weld defect detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN116823756A true CN116823756A (en) | 2023-09-29 |
Family
ID=88121547
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310745818.5A Pending CN116823756A (en) | 2023-06-25 | 2023-06-25 | Pile leg weld defect detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116823756A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117252883A (en) * | 2023-11-20 | 2023-12-19 | 宁德时代新能源科技股份有限公司 | Welding defect detection method, device and storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014139273A1 (en) * | 2013-03-14 | 2014-09-18 | 苏州华源包装股份有限公司 | Weld seam defect detection method |
CN109636772A (en) * | 2018-10-25 | 2019-04-16 | 同济大学 | The defect inspection method on the irregular shape intermetallic composite coating surface based on deep learning |
CN111932489A (en) * | 2020-06-03 | 2020-11-13 | 西安电子科技大学 | Weld defect detection method, system, storage medium, computer device and terminal |
WO2020248439A1 (en) * | 2019-06-11 | 2020-12-17 | 江苏农林职业技术学院 | Crown cap surface defect online inspection method employing image processing |
CN113379703A (en) * | 2021-06-08 | 2021-09-10 | 西安理工大学 | Photovoltaic panel dark spot defect detection method based on Yolo-v4 network structure |
CN116309277A (en) * | 2022-12-14 | 2023-06-23 | 郑州金惠计算机系统工程有限公司 | Steel detection method and system based on deep learning |
-
2023
- 2023-06-25 CN CN202310745818.5A patent/CN116823756A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014139273A1 (en) * | 2013-03-14 | 2014-09-18 | 苏州华源包装股份有限公司 | Weld seam defect detection method |
CN109636772A (en) * | 2018-10-25 | 2019-04-16 | 同济大学 | The defect inspection method on the irregular shape intermetallic composite coating surface based on deep learning |
WO2020248439A1 (en) * | 2019-06-11 | 2020-12-17 | 江苏农林职业技术学院 | Crown cap surface defect online inspection method employing image processing |
CN111932489A (en) * | 2020-06-03 | 2020-11-13 | 西安电子科技大学 | Weld defect detection method, system, storage medium, computer device and terminal |
CN113379703A (en) * | 2021-06-08 | 2021-09-10 | 西安理工大学 | Photovoltaic panel dark spot defect detection method based on Yolo-v4 network structure |
CN116309277A (en) * | 2022-12-14 | 2023-06-23 | 郑州金惠计算机系统工程有限公司 | Steel detection method and system based on deep learning |
Non-Patent Citations (1)
Title |
---|
张晓光, 林家骏: "X射线检测焊缝的图像处理与缺陷识别", 华东理工大学学报, no. 02, 30 April 2004 (2004-04-30) * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN117252883A (en) * | 2023-11-20 | 2023-12-19 | 宁德时代新能源科技股份有限公司 | Welding defect detection method, device and storage medium |
CN117252883B (en) * | 2023-11-20 | 2024-04-02 | 宁德时代新能源科技股份有限公司 | Welding defect detection method, device and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108961235B (en) | Defective insulator identification method based on YOLOv3 network and particle filter algorithm | |
CN107230202B (en) | Automatic identification method and system for road surface disease image | |
CN110033471B (en) | Frame line detection method based on connected domain analysis and morphological operation | |
CN103020605B (en) | Bridge identification method based on decision-making layer fusion | |
CN109145872B (en) | CFAR and Fast-RCNN fusion-based SAR image ship target detection method | |
CN102750540B (en) | Morphological filtering enhancement-based maximally stable extremal region (MSER) video text detection method | |
CN113128442B (en) | Chinese character handwriting style identification method and scoring method based on convolutional neural network | |
CN107301405A (en) | Method for traffic sign detection under natural scene | |
CN107680090A (en) | Based on the electric transmission line isolator state identification method for improving full convolutional neural networks | |
CN111833328B (en) | Aircraft engine blade surface defect detection method based on deep learning | |
CN111611907B (en) | Image-enhanced infrared target detection method | |
CN107038416B (en) | Pedestrian detection method based on binary image improved HOG characteristics | |
CN107506765B (en) | License plate inclination correction method based on neural network | |
CN109840483B (en) | Landslide crack detection and identification method and device | |
CN104182985A (en) | Remote sensing image change detection method | |
CN111666842A (en) | Shadow detection method based on double-current-cavity convolution neural network | |
CN116823756A (en) | Pile leg weld defect detection method | |
CN117036737A (en) | Feature extraction and matching method based on information entropy, GMS and LC significant detection | |
CN108932471B (en) | Vehicle detection method | |
CN114429577A (en) | Flag detection method, system and equipment based on high beacon strategy | |
CN112396582B (en) | Mask RCNN-based equalizing ring skew detection method | |
CN113095283A (en) | Lane line extraction method based on dynamic ROI and improved firefly algorithm | |
CN109829511B (en) | Texture classification-based method for detecting cloud layer area in downward-looking infrared image | |
CN111709910A (en) | PCB defect detection algorithm based on convolutional neural network | |
CN116363655A (en) | Financial bill identification method and system |
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 | ||
CB02 | Change of applicant information |
Address after: 214200 No.16 Xingli Road, Yixing Economic and Technological Development Zone, Wuxi City, Jiangsu Province Applicant after: Jiangsu Yongkang Intelligent Defense Technology Co.,Ltd. Address before: 214200 No.16 Xingli Road, Yixing Economic and Technological Development Zone, Wuxi City, Jiangsu Province Applicant before: JIANGSU YONGKANG MACHINERY Co.,Ltd. |
|
CB02 | Change of applicant information |