CN117911411B - Computer vision detection method and device based on parallel detection of picture streams - Google Patents
Computer vision detection method and device based on parallel detection of picture streams Download PDFInfo
- Publication number
- CN117911411B CN117911411B CN202410312766.7A CN202410312766A CN117911411B CN 117911411 B CN117911411 B CN 117911411B CN 202410312766 A CN202410312766 A CN 202410312766A CN 117911411 B CN117911411 B CN 117911411B
- Authority
- CN
- China
- Prior art keywords
- product
- picture
- local
- local picture
- detection
- 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.)
- Active
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 115
- 230000007547 defect Effects 0.000 claims abstract description 17
- 238000004519 manufacturing process Methods 0.000 claims abstract description 12
- 238000000034 method Methods 0.000 claims abstract description 7
- 238000007689 inspection Methods 0.000 claims abstract description 6
- 238000004458 analytical method Methods 0.000 claims description 17
- 230000010354 integration Effects 0.000 claims description 5
- 238000005516 engineering process Methods 0.000 description 7
- 230000004075 alteration Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013136 deep learning model Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 239000007943 implant Substances 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
- 230000000007 visual effect Effects 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 Processing (AREA)
Abstract
The invention discloses a computer vision detection method and device based on parallel detection of picture streams. According to the method, a plurality of array surface cameras are adopted to shoot a product moving on a production line for multiple times, each array surface camera shoots and acquires a local picture of the product each time, then the local picture is encoded, a code obtained by encoding is implanted into the name of the local picture of the product, an event to be detected is generated according to the name of the local picture of the product, a corresponding part detection model is called in parallel according to the name of the local picture of the product in the event to be detected to detect defects of the local picture of the product, a detection result of each local picture is obtained, the detection results of the local pictures belonging to the same product are integrated, and the number of detected parts is counted. The invention realizes shooting and detection simultaneously, improves the speed and efficiency of product quality inspection, meets the requirement on real-time performance of industrial product online detection, and reduces project implementation cost.
Description
Technical Field
The invention relates to the technical field of product defect detection, in particular to a computer vision detection method and device based on parallel detection of picture streams.
Background
At present, in the process of advancing the digital upgrading of industrial manufacturing and implementing intelligent manufacturing, one of the key problems is how to detect the appearance quality of industrial products manufactured on a production line by using an automatic technology, thereby realizing an end-to-end automatic production line from raw material processing and production to rear end appearance quality detection. The computer vision detection is an effective technology for automatically detecting the appearance quality of industrial products, which is widely applied in practical engineering, and the basic idea of the technology is that firstly, the industrial products to be detected are photographed through an industrial camera, the photographed pictures are sent to a detection model, such as a deep learning model, for detection analysis, and finally, the detection result is fed back to an actual user, or the industrial products to be detected are sorted according to the detection result.
The size of the sample is different for different industrial products, some of the industrial products are relatively small, such as buttons, and some of the industrial products are relatively large, such as LCD TV display screens. Because of the limitation of the shooting visual field of the industrial camera, when an area array camera is adopted for large-size industrial products to be detected, one shot photo cannot completely cover the whole industrial products to be detected. In this case, conventionally, photographing and detection are often performed in two ways.
In a first mode, a plurality of area array cameras are adopted to shoot an industrial product to be detected from different angles at the same time, each photo covers a certain specific part of the industrial product to be detected, a plurality of photos are combined to form a complete picture of the product, and a detection model is called to detect the picture based on the complete picture of the product. In this way, since a plurality of cameras are required to photograph the industrial product to be inspected at the same time, a plurality of cameras are required to be purchased, and the investment in engineering is relatively high.
And secondly, scanning by adopting a line scanning camera mode in a mode of moving the camera or moving the industrial product to be detected, outputting a complete picture of the whole product after the scanning is finished, and then calling a detection model to analyze the picture to obtain an analysis result. In this way, the detection is performed after waiting for the movement of the camera or the picture to scan the whole picture, and meanwhile, the whole picture of the whole product is relatively large, so that the detection time is also long, and therefore, the real-time detection is often difficult to achieve by the detection way.
Disclosure of Invention
The invention aims at overcoming the defects of the prior art and provides a computer vision detection method and device based on parallel detection of picture streams.
To achieve the above object, in a first aspect, a computer vision detection method based on parallel detection of picture streams includes:
Shooting a product moving on a production line for multiple times by adopting a plurality of array face cameras, wherein each array face camera shoots and acquires a local picture of the product each time, coding the local picture of the product according to different array face cameras and shooting orders thereof, then implanting the code obtained by coding into the name of the local picture of the product, generating an event to be detected according to the name of the local picture of the product, and caching the local picture of the product and the event to be detected;
Constructing a corresponding part detection model according to the parts of the product, and storing the part detection model;
Obtaining a cached local picture of a product and an event to be detected of the local picture, calling a corresponding component detection model to detect defects of the local picture of the product according to the name of the local picture of the product in the event to be detected, generating a detection result for the local picture of each product, and caching the detection result of the local picture of each product;
And acquiring a detection result of each local picture, integrating the detection results of the local pictures belonging to the same product, counting the number of detected parts, judging that the product is qualified if the detection result of each local picture is qualified and the number of the detected parts is the same as the set number of the parts, and judging that the product is unqualified if the product is not qualified.
Further, the numbers include a partial picture number+a camera number.
Further, the local picture number is obtained in the following manner:
The local picture number of the first shot product is set to 01 for the initial part picture, 10 for the last shot end part picture, and 00 for the middle part picture.
Further, the manner of numbering the cameras is specifically as follows:
two-bit numbers mn are adopted to represent an mn-number array surface camera, and m and n are natural numbers.
Further, the picture number and the camera number are set at the end of the name of the partial picture of the product.
In a second aspect, the present invention provides a computer vision detection device based on parallel detection of picture streams, comprising:
The system comprises a plurality of array surface cameras, a plurality of camera modules and a plurality of camera modules, wherein the array surface cameras are used for shooting a product moving on a production line for a plurality of times, and each array surface camera shoots and collects a local picture of the product each time;
The coding module is used for coding the local pictures of the product according to different array cameras and shooting orders thereof, then implanting the codes obtained by coding into the names of the local pictures of the product, and generating an event to be detected according to the names of the local pictures of the product;
The caching module is used for caching the local pictures of the product and the events to be detected of the local pictures;
The model analysis service module is used for constructing a corresponding part detection model according to the parts of the product and storing the part detection model;
The model analysis parallel scheduling module is used for acquiring the cached local pictures of the products and events to be detected, and concurrently calling a corresponding component detection model to detect the defects of the local pictures of the products according to the names of the local pictures of the products in the events to be detected, and the model analysis service module is also used for generating detection results for the local pictures of each product;
the model analysis result caching module is used for caching the detection result of the local picture of each product;
The result integration module is used for acquiring the detection result of each local picture, integrating the detection results of the local pictures belonging to the same product, counting the number of detected parts, judging that the product is detected to be qualified if the detection result of each local picture is qualified and the number of the detected parts is the same as the set number of the parts, and judging that the product is detected to be unqualified if the detection result of each local picture is not qualified.
Further, the numbers include a partial picture number+a camera number.
Further, the local picture number is obtained in the following manner:
For the first shot starting part picture of the product, the local picture number is set to 01, for the last shot ending part picture of the product, the local picture number is set to 10, and for the middle part picture, the local picture number is set to 00.
Further, the manner of numbering the cameras is specifically as follows:
two-bit numbers mn are adopted to represent an mn-number array surface camera, and m and n are natural numbers.
Further, the picture number and the camera number are set at the end of the name of the partial picture of the product.
The beneficial effects are that: 1. according to the method, the local pictures of the product are collected and encoded, then the event to be detected is generated, the component detection model is called based on the event to be detected to carry out parallel detection, so that the detection is realized while shooting, the speed and the efficiency of quality inspection of the product are improved, and the requirement on real-time performance of industrial product online detection is met;
2. Based on the mode of shooting and detecting simultaneously, shooting of different parts of a product by the same camera is achieved, investment on the number of cameras in engineering is reduced, and project implementation cost is reduced.
Drawings
Fig. 1 is a schematic block diagram of a computer vision inspection apparatus based on parallel inspection of picture streams in accordance with an embodiment of the present invention.
Detailed Description
The invention will be further illustrated by the following drawings and specific examples, which are carried out on the basis of the technical solutions of the invention, it being understood that these examples are only intended to illustrate the invention and are not intended to limit the scope of the invention.
The embodiment of the invention provides a computer vision detection method based on parallel detection of picture streams, which comprises the following steps:
And shooting the moving product on the production line for a plurality of times by adopting a plurality of array face cameras, wherein each array face camera shoots and acquires a local picture of the product each time, coding the local picture of the product according to different array face cameras and shooting orders thereof, then implanting the code obtained by coding into the name of the local picture of the product, generating an event to be detected according to the name of the local picture of the product, and caching the local picture of the product and the event to be detected. Therefore, the picture information of different parts of the product can be distinguished according to the preset parameter information through the names of the local pictures.
Specifically, the numbers include a partial picture number+a camera number. The local picture numbers are obtained in the following manner:
The local picture number of the first shot product is set to 01 for the initial part picture, 10 for the last shot end part picture, and 00 for the middle part picture. In addition, when the product enters the shooting area, the image can be detected through triggering of an infrared sensor or an image recognition technology, then an acquisition starting instruction is generated, the first acquired image after the array camera receives the acquisition starting instruction is the initial position image, when the product leaves the shooting area, the image can also be detected through triggering of the infrared sensor or the image recognition technology, then an acquisition stopping instruction is generated, the image acquired before the array camera receives the acquisition stopping instruction is the end position image, and the image acquired between the initial position image and the end position image is the middle part image.
The camera number is obtained in the following manner:
two-bit numbers mn are adopted to represent an mn-number array surface camera, and m and n are natural numbers. Specifically, it is preferable to start with 01, number 01 denotes an array camera No. 1, number 02 denotes an array camera No. 2, and so on. The above-mentioned picture number and camera number are preferably set at the end of the name of the local picture of the product, in particular the last four digits of the name of the local picture of the product. In addition, for one product, the above-mentioned middle part picture may be plural, and although the last four bits of the names of the partial pictures are the same, the top N bits in the names of the partial pictures are different, and thus can be distinguished according to the top N bits of the names of the partial pictures. The first N bits of the name of the partial picture may represent the shooting order by taking the shooting time or other means.
And constructing a corresponding part detection model according to the parts of the product, and storing the part detection model. Each component detection model may be used to identify and detect a component in the acquired partial picture.
Obtaining a cached local picture of a product and an event to be detected of the local picture, calling a corresponding component detection model to detect defects of the local picture of the product according to the name of the local picture of the product in the event to be detected, generating a detection result for the local picture of each product, and caching the detection result of the local picture of each product, wherein the detection result comprises the identified component name and whether the identified component name is qualified or not. It should be noted that, the area of the product collected by the initial part picture, each middle part picture, and the final part picture is set in advance, so that according to the name of the local picture of the product in the event to be detected, it can be determined which parts of the product are included in the picture, and then it can be determined which part detection model needs to be called for defect detection for the local picture. The following two layers of detection can be completed through the local pictures of the single product: 1. a defect in a localized portion of the product itself; 2. the distance between two different parts of the product is not satisfactory. For defects of some parts of the product, which are not right in number, the defects cannot be detected, so that the detection result of each partial picture of the product needs to be integrated and analyzed.
And acquiring a detection result of each local picture, integrating the detection results of the local pictures belonging to the same product, counting the number of detected parts, judging that the product is qualified if the detection result of each local picture is qualified and the number of the detected parts is the same as the set number of the parts, and judging that the product is unqualified if the detection result of any local picture is unqualified or the number of the detected parts is different from the set number of the parts. The set number of parts is the actual number of parts of the product. Judging whether the local pictures belong to the same product or not can be analyzed through the names of the local pictures, specifically, a certain initial part picture, an end part picture acquired for the first time later and a middle part picture between the initial part picture and the end part picture belong to the local pictures of the same product, and the local pictures can be used as the basis of the same product for detection result integration.
Referring to fig. 1, based on the above embodiments, it can be easily understood by those skilled in the art that the present invention further provides a computer vision detection device based on parallel detection of picture streams, which includes a plurality of array camera 1, a coding module 2, a buffer module 3, a model analysis service module 4, a model analysis parallel scheduling module 5, a model analysis result buffer module 6 and a result integration module 7.
The array camera 1 is used for shooting a product moving on a production line for a plurality of times, and each array camera 1 shoots and collects a local picture of the product each time.
The encoding module 2 is configured to encode the local pictures of the product according to different cameras and shooting orders thereof, then implant the numbers obtained by encoding into the names of the local pictures of the product, and generate a to-be-detected event according to the names of the local pictures of the product. Therefore, the picture information of different parts of the product can be distinguished according to the preset parameter information through the names of the local pictures.
Specifically, the numbers include a partial picture number+a camera number. The local picture numbers are obtained in the following manner:
The local picture number of the first shot product is set to 01 for the initial part picture, 10 for the last shot end part picture, and 00 for the middle part picture. In addition, when the product enters the shooting area, the image can be detected through triggering of an infrared sensor or an image recognition technology, then an acquisition starting instruction is generated, the first acquired image after the array camera receives the acquisition starting instruction is the initial position image, when the product leaves the shooting area, the image can also be detected through triggering of the infrared sensor or the image recognition technology, then an acquisition stopping instruction is generated, the image acquired before the array camera receives the acquisition stopping instruction is the end position image, and the image acquired between the initial position image and the end position image is the middle part image.
The camera number is obtained in the following manner:
two-bit numbers mn are adopted to represent an mn-number array surface camera, and m and n are natural numbers. Specifically, it is preferable to start with 01, number 01 denotes an array camera No. 1, number 02 denotes an array camera No. 2, and so on. The above-mentioned picture number and camera number are preferably set at the end of the name of the local picture of the product, in particular the last four digits of the name of the local picture of the product. In addition, for one product, the above-mentioned middle part picture may be plural, and although the last four bits of the names of the partial pictures are the same, the top N bits in the names of the partial pictures are different, and thus can be distinguished according to the top N bits of the names of the partial pictures. The first N bits of the name of the partial picture may represent the shooting order by taking the shooting time or other means.
The buffer module 3 is configured to buffer the local image of the product acquired by the array camera 1 and the event to be detected generated by the encoding module 2.
The model analysis service module 4 is configured to construct a corresponding component detection model according to the component of the product, and store the component detection model. Each component detection model may be used to identify and detect a component in the acquired partial picture.
The model analysis parallel scheduling module 5 is configured to obtain a cached local picture of a product and an event to be detected thereof, and to call a corresponding component detection model to detect a defect of the local picture of the product according to the name of the local picture of the product in the event to be detected, and the model analysis service module 4 is further configured to generate a detection result for the local picture of each product, where the detection result includes the identified component name and whether the identified component name is qualified or not. It should be noted that, the area of the product collected by the initial part picture, each middle part picture, and the final part picture is set in advance, so that according to the name of the local picture of the product in the event to be detected, it can be determined which parts of the product are included in the picture, and then it can be determined which part detection model needs to be called for defect detection for the local picture. The following two layers of detection can be completed through the local pictures of the single product: 1. a defect in a localized portion of the product itself; 2. the distance between two different parts of the product is not satisfactory. For defects of some parts of the product, which are not right in number, the defects cannot be detected, so that the detection result of each partial picture of the product needs to be integrated and analyzed.
The model analysis result caching module 6 is configured to cache the detection result of the local picture of each product.
The result integrating module 7 is configured to obtain a detection result of each local picture, integrate the detection results of the local pictures belonging to the same product, and count the number of detected parts, if the detection result of each local picture is qualified, and the number of detected parts is the same as the set number of parts, determine that the product is qualified for detection, and if the detection result of any local picture is unqualified, or if the number of detected parts is different from the set number of parts, determine that the product is unqualified for detection. The set number of parts is the actual number of parts of the product. Judging whether the local pictures belong to the same product or not can be analyzed through the names of the local pictures, specifically, a certain initial part picture, an end part picture acquired for the first time later and a middle part picture between the initial part picture and the end part picture belong to the local pictures of the same product, and the local pictures can be used as the basis of the same product for detection result integration.
The foregoing is merely a preferred embodiment of the present invention, and it should be noted that other parts not specifically described are within the prior art or common general knowledge to a person of ordinary skill in the art. Modifications and alterations may be made without departing from the principles of this invention, and such modifications and alterations should also be considered as being within the scope of the invention.
Claims (8)
1. A computer vision detection method based on parallel detection of picture streams is characterized by comprising the following steps:
Shooting a product moving on a production line for multiple times by adopting a plurality of array face cameras, wherein each array face camera shoots and acquires a local picture of a set area of the product each time, coding the local picture of the product according to different array face cameras and shooting orders thereof, then implanting a code obtained by coding into the name of the local picture of the product, generating an event to be detected according to the name of the local picture of the product, and caching the local picture of the product and the event to be detected, wherein the code comprises a local picture number and a camera number;
Constructing a corresponding part detection model according to the parts of the product, and storing the part detection model;
Obtaining a local picture of a cached product and an event to be detected of the local picture, determining a part of the product contained in the local picture according to the name of the local picture of the product in the event to be detected, concurrently calling a corresponding part detection model according to the part of the product contained in the local picture to detect defects of the local picture of the product, generating a detection result for the local picture of each product, and caching the detection result of the local picture of each product;
And acquiring a detection result of each local picture, integrating the detection results of the local pictures belonging to the same product, counting the number of detected parts, judging that the product is qualified if the detection result of each local picture is qualified and the number of the detected parts is the same as the set number of the parts, and judging that the product is unqualified if the product is not qualified.
2. The method for detecting computer vision based on parallel detection of picture streams according to claim 1, wherein the local picture numbers are obtained by the following steps:
The local picture number of the first shot product is set to 01 for the initial part picture, 10 for the last shot end part picture, and 00 for the middle part picture.
3. The method for detecting computer vision based on parallel detection of picture streams according to claim 1, wherein the camera number is obtained by the following specific ways:
two-bit numbers mn are adopted to represent an mn-number array surface camera, and m and n are natural numbers.
4. The method for detecting the local picture according to claim 1, wherein the local picture number and the camera number are set at the end of the name of the local picture of the product.
5. A computer vision inspection device based on parallel inspection of picture streams, comprising:
the array surface cameras are used for shooting the product moving on the production line for multiple times, and each array surface camera shoots and collects local pictures of the product set area each time;
The coding module is used for coding the local pictures of the product according to different array cameras and shooting orders thereof, then implanting the codes obtained by coding into the names of the local pictures of the product, and generating an event to be detected according to the names of the local pictures of the product, wherein the codes comprise the local picture numbers and the camera numbers;
The caching module is used for caching the local pictures of the product and the events to be detected of the local pictures;
The model analysis service module is used for constructing a corresponding part detection model according to the parts of the product and storing the part detection model;
The model analysis parallel scheduling module is used for acquiring a cached local picture of a product and an event to be detected of the local picture, determining a part of the product contained in the local picture according to the name of the local picture of the product in the event to be detected, and carrying out defect detection on the local picture of the product according to a corresponding part detection model which is concurrently called by the part of the product contained in the local picture, wherein the model analysis service module is also used for generating a detection result for the local picture of each product;
the model analysis result caching module is used for caching the detection result of the local picture of each product;
The result integration module is used for acquiring the detection result of each local picture, integrating the detection results of the local pictures belonging to the same product, counting the number of detected parts, judging that the product is detected to be qualified if the detection result of each local picture is qualified and the number of the detected parts is the same as the set number of the parts, and judging that the product is detected to be unqualified if the detection result of each local picture is not qualified.
6. The device for detecting computer vision based on parallel detection of picture streams according to claim 5, wherein the local picture numbers are obtained by the following steps:
The local picture number of the first shot product is set to 01 for the initial part picture, 10 for the last shot end part picture, and 00 for the middle part picture.
7. The apparatus for detecting computer vision based on parallel detection of picture streams according to claim 5, wherein the camera number is obtained by the following method:
two-bit numbers mn are adopted to represent an mn-number array surface camera, and m and n are natural numbers.
8. The apparatus of claim 5, wherein the local picture number and the camera number are set at the end of the name of the local picture of the product.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410312766.7A CN117911411B (en) | 2024-03-19 | 2024-03-19 | Computer vision detection method and device based on parallel detection of picture streams |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410312766.7A CN117911411B (en) | 2024-03-19 | 2024-03-19 | Computer vision detection method and device based on parallel detection of picture streams |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117911411A CN117911411A (en) | 2024-04-19 |
| CN117911411B true CN117911411B (en) | 2024-05-24 |
Family
ID=90684004
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410312766.7A Active CN117911411B (en) | 2024-03-19 | 2024-03-19 | Computer vision detection method and device based on parallel detection of picture streams |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117911411B (en) |
Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104458749A (en) * | 2013-09-25 | 2015-03-25 | 中国科学院沈阳自动化研究所 | Real-time aluminum profile surface defect detection system based on machine vision |
| CN104990512A (en) * | 2015-07-20 | 2015-10-21 | 清华大学 | Transparent container bottle opening thread defect detection system and method |
| CN108109137A (en) * | 2017-12-13 | 2018-06-01 | 重庆越畅汽车科技有限公司 | The Machine Vision Inspecting System and method of vehicle part |
| CN209745832U (en) * | 2018-12-07 | 2019-12-06 | 北京黎马敦太平洋包装有限公司 | printed product detection equipment that single-camera multi-angle detected |
| CN110567977A (en) * | 2019-10-11 | 2019-12-13 | 湖南讯目科技有限公司 | Curved glass defect detection system and method |
| CN112964724A (en) * | 2021-02-01 | 2021-06-15 | 苏州百迈半导体技术有限公司 | Multi-target multi-zone visual detection method and system |
| CN113205136A (en) * | 2021-04-30 | 2021-08-03 | 浙大城市学院 | Real-time high-precision detection method for appearance defects of power adapter |
| WO2022204788A1 (en) * | 2021-03-29 | 2022-10-06 | Musashi Auto Parts Canada Inc. | System, method, and computer device for automated visual inspection using adaptive region of interest segmentation |
| CN115423785A (en) * | 2022-09-14 | 2022-12-02 | 杭州海康机器人股份有限公司 | Defect detection system, method and device, electronic equipment and storage medium |
| CN115983687A (en) * | 2022-12-22 | 2023-04-18 | 广州海秀敏网络科技有限公司 | Intelligent detection management system and method for quality of cold-rolled strip steel |
| CN116908106A (en) * | 2023-08-08 | 2023-10-20 | 潍柴动力股份有限公司 | Product detection method and device based on vision |
| CN116994024A (en) * | 2022-11-30 | 2023-11-03 | 腾讯科技（深圳）有限公司 | Method, device, equipment, medium and product for identifying parts in container image |
| CN117197090A (en) * | 2023-09-12 | 2023-12-08 | 信华信(大连)软件服务股份有限公司 | Disposable tableware surface defect detection system based on deep learning |
| CN117230567A (en) * | 2023-09-18 | 2023-12-15 | 福建省金奥智能装备有限公司 | Winding and cloth inspecting method of circular knitting machine |
| CN117475205A (en) * | 2023-10-23 | 2024-01-30 | 广西广盛新材料科技有限公司 | Defect type identification method, defect type identification device, control device and storage medium |
-
2024
- 2024-03-19 CN CN202410312766.7A patent/CN117911411B/en active Active
Patent Citations (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104458749A (en) * | 2013-09-25 | 2015-03-25 | 中国科学院沈阳自动化研究所 | Real-time aluminum profile surface defect detection system based on machine vision |
| CN104990512A (en) * | 2015-07-20 | 2015-10-21 | 清华大学 | Transparent container bottle opening thread defect detection system and method |
| CN108109137A (en) * | 2017-12-13 | 2018-06-01 | 重庆越畅汽车科技有限公司 | The Machine Vision Inspecting System and method of vehicle part |
| CN209745832U (en) * | 2018-12-07 | 2019-12-06 | 北京黎马敦太平洋包装有限公司 | printed product detection equipment that single-camera multi-angle detected |
| CN110567977A (en) * | 2019-10-11 | 2019-12-13 | 湖南讯目科技有限公司 | Curved glass defect detection system and method |
| CN112964724A (en) * | 2021-02-01 | 2021-06-15 | 苏州百迈半导体技术有限公司 | Multi-target multi-zone visual detection method and system |
| WO2022204788A1 (en) * | 2021-03-29 | 2022-10-06 | Musashi Auto Parts Canada Inc. | System, method, and computer device for automated visual inspection using adaptive region of interest segmentation |
| CN113205136A (en) * | 2021-04-30 | 2021-08-03 | 浙大城市学院 | Real-time high-precision detection method for appearance defects of power adapter |
| CN115423785A (en) * | 2022-09-14 | 2022-12-02 | 杭州海康机器人股份有限公司 | Defect detection system, method and device, electronic equipment and storage medium |
| CN116994024A (en) * | 2022-11-30 | 2023-11-03 | 腾讯科技（深圳）有限公司 | Method, device, equipment, medium and product for identifying parts in container image |
| CN115983687A (en) * | 2022-12-22 | 2023-04-18 | 广州海秀敏网络科技有限公司 | Intelligent detection management system and method for quality of cold-rolled strip steel |
| CN116908106A (en) * | 2023-08-08 | 2023-10-20 | 潍柴动力股份有限公司 | Product detection method and device based on vision |
| CN117197090A (en) * | 2023-09-12 | 2023-12-08 | 信华信(大连)软件服务股份有限公司 | Disposable tableware surface defect detection system based on deep learning |
| CN117230567A (en) * | 2023-09-18 | 2023-12-15 | 福建省金奥智能装备有限公司 | Winding and cloth inspecting method of circular knitting machine |
| CN117475205A (en) * | 2023-10-23 | 2024-01-30 | 广西广盛新材料科技有限公司 | Defect type identification method, defect type identification device, control device and storage medium |
Non-Patent Citations (3)
| Title |
|---|
| Multi-Camera-Based Sorting System for Surface Defects of Apples;Ju-Hwan Lee等;《sensors》;20230413;第23卷;1-20 * |
| 基于机器视觉技术的钢板表面质量检测系统;高冰等;《第十一届中国钢铁年会论文集——S18.冶金自动化与智能管控》;20171231;1-5 * |
| 多工位硅胶圈缺陷在线视觉检测系统研究;毕德学等;《机械设计与制造》;20240229(第2期);125-130 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117911411A (en) | 2024-04-19 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102630299B (en) | Image processing device for defect inspection and image processing method for defect inspection | |
| CN105812746B (en) | A kind of object detection method and system | |
| JPH04166751A (en) | Method and apparatus for inspecting defect in bottle and the like | |
| CN110599445A (en) | Target robust detection and defect identification method and device for power grid nut and pin | |
| CN101621628A (en) | Photographic apparatus, setting method of photography conditions, and recording medium | |
| JP2003244521A (en) | Information processing method and apparatus, and recording medium | |
| CN107833843A (en) | The analysis method and analysis system of defect source, defect detecting device | |
| CN107959841A (en) | Image processing method, device, storage medium and electronic equipment | |
| CN114078127B (en) | Object defect detection and counting method, device, equipment and storage medium | |
| CN117911411B (en) | Computer vision detection method and device based on parallel detection of picture streams | |
| CN105844282A (en) | Method for detecting defects of fuel injection nozzle O-Ring through line scanning camera | |
| CN116580026B (en) | Automatic optical detection method, equipment and storage medium for appearance defects of precision parts | |
| CN111247790A (en) | Image processing method and device, image shooting and processing system and carrier | |
| CN109063562A (en) | Image processing method and device, electronic equipment, computer readable storage medium | |
| CN1018099B (en) | Determine the method and apparatus of number in space specified | |
| JPH08210847A (en) | Image processing method | |
| CN101841654A (en) | Image processing apparatus and image processing method | |
| CN115272340B (en) | Industrial product defect detection method and device | |
| CN116091506A (en) | Machine vision defect quality inspection method based on YOLOV5 | |
| KR100649384B1 (en) | Method for extracting image characteristic in low intensity of illumination environment | |
| EP0209422A1 (en) | Automatic digitalisation methods and apparatuses for scenes comprising discrete significant elements | |
| CN114049317A (en) | CCD equipment automatic inspection system and method based on artificial intelligence | |
| CN108898163A (en) | Information processing method and device, electronic equipment and computer readable storage medium | |
| CN113824874A (en) | Auxiliary shooting method and device, electronic equipment and storage medium | |
| CN114692663B (en) | Photographing identification fault tolerance method for code reading failure |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |