CN111259184A - Image automatic labeling system and method for new retail - Google Patents
Image automatic labeling system and method for new retail Download PDFInfo
- Publication number
- CN111259184A CN111259184A CN202010122977.6A CN202010122977A CN111259184A CN 111259184 A CN111259184 A CN 111259184A CN 202010122977 A CN202010122977 A CN 202010122977A CN 111259184 A CN111259184 A CN 111259184A
- Authority
- CN
- China
- Prior art keywords
- picture
- target
- marked
- position information
- label
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/5866—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
Abstract
The invention relates to an automatic image annotation system and method for new retail. The embodiment comprises the steps of obtaining a picture to be marked and a test unit which is corresponding to the picture to be marked and is used for automatically marking the picture, obtaining position information of a target in the picture to be marked and a cascade network test unit of a label of the target, checking whether a marking result is a test unit of group route, and obtaining the test unit of the marking result. Training an image segmentation model for a target to be labeled, shooting a picture to be labeled, placing a picture for automatic labeling with a digital paper piece on the target of the picture to be labeled, acquiring position information of the target in the picture through an image segmentation network, checking whether the target is wrong, acquiring digital information on the target through a deep learning network, confirming a label of the target through a corresponding relation between a number and the label after checking that the target is correct, and generating a csv format file by using the position information and the label of each target in the picture to finish automatic labeling.
Description
Technical Field
The invention relates to the technical field of image annotation, in particular to an automatic image annotation system and method for new retail.
Background
Image annotation is a technique for noting the position and label of an object in an image. Due to the popularization of the artificial intelligence technology, data annotation appears in the front of people as a new industry, in the field of machine vision, image annotation becomes a complex and tedious work with a large demand, and the improvement of the efficiency of image annotation becomes a hotspot.
In a conventional labeling method, a target object position is marked in an original image, and corresponding label file indicating position information of a target and a label of the target are generated for each picture. The traditional labeling method has high reliability, but has the problems of low efficiency and serious waste of human resources. The existing automatic labeling technology is to label by using a trained target identification network aiming at a specific scene, and the labeling method has high automation degree, but can not ensure the correctness of the labeling result, and has strong pertinence and poor expansibility of the label. Compared with the prior art, the automatic image labeling method has the advantages that the labeling process is automatic labeling, a large batch of pictures can be rapidly processed, the efficiency is high, the accuracy of labeling is ensured by artificially checking the accuracy of labeling in a key link, the first network realizes image segmentation and the second network completes the task of digital identification by utilizing a cascaded network structure, and the flexibility of adding labels is enhanced by a method of corresponding the numbers to the labels.
Disclosure of Invention
The invention aims to provide an automatic image annotation method for new retail so as to overcome the defects in the prior art.
In order to achieve the purpose, the technical scheme of the invention is as follows: an automatic image annotation system for new retail sales, comprising:
the device comprises a test unit, a platform, a camera, a target and a test unit, wherein the test unit is used for obtaining a picture to be marked and the picture to be marked, which corresponds to the picture to be marked and is used for automatically marking the picture;
the cascade network testing unit is used for acquiring the position information of a target in a picture to be marked and a label of the target, acquiring the position information of the target in the picture through a trained image segmentation network aiming at the target, identifying numbers placed on the target through a deep learning network, and acquiring the label of the target through the corresponding relation between the numbers and the label;
the test unit is used for checking whether the marking result is a group Truth result or not, and is used for confirming whether the marking result is accurate or not, checking whether the position information has errors or not when the position information in the picture to be marked is acquired by using an image segmentation network, selecting the wrong picture for marking if the position information has errors, displaying each target and the number corresponding to each target after the number on the target is identified, detecting whether the number has errors or not, and modifying the label of the target if the number has errors;
and the test unit is used for generating a csv file, and writing the position information of each target of each picture, the label of the target and the name of the picture into the csv format file through the written python code and storing the csv format file.
The invention also provides a new retail image automatic labeling method, which comprises the following steps:
step S1: establishing a data set for a target to be marked, and training an image segmentation model through the data set to acquire position information of the target to be marked;
step S2: through a device with a fixed relative position between a platform and a camera, after a picture to be marked is shot, a paper sheet with numbers is placed on a target, and the picture to be marked is shot again to obtain a corresponding picture for automatic marking of the picture to be marked, so that a data set to be marked is manufactured;
step S3: obtaining the position information of each target in the corresponding picture for automatic annotation of the annotated picture through the image segmentation model in the step S1, checking whether the position information is wrong, and if the position information is wrong, continuing to annotate the annotated picture and saving the annotated picture in a folder for saving the picture to be annotated;
step S4: intercepting the image of each target in the image according to the position information of each target, identifying the number on the target through a deep learning network, and confirming the label of the target through the corresponding relation between the number and the label after checking that no error exists;
step S5: and compiling python codes, and making the file name of each picture, the position information and the label of each target in the picture into a csv format file to finish automatic labeling.
In an embodiment of the present invention, in the step S2, the method further includes the following steps:
step S201: two folders are established for each picture to be marked, one folder stores the picture to be marked, the other folder stores the picture of the paper sheet with the number written on the target, the two pictures are shot by equipment with fixed relative positions of a camera and a platform for placing the target, and the names of the two pictures are the same.
In an embodiment of the present invention, in the step S3, the method further includes the following steps:
step S301: and adding a rectangular frame to each target according to the position information of each target, displaying the framed picture, storing the picture to be detected into a folder for storing the picture to be marked if the picture is checked to be correct, and performing the next step if the picture is correct.
In an embodiment of the present invention, in the step S4, the method further includes the following steps:
step S401: intercepting pictures of each target in the targets according to the position information of each target, identifying the numbers of the cards with the numbers written on the targets through a trained deep learning network, and labeling the targets in the area according to the corresponding relation between the numbers and the labels;
step S402: during checking, the picture of each target intercepted in the step S401 in the picture is displayed, the recognized number is displayed at the upper left corner of the picture of each target, and the next step is carried out after checking is correct.
Compared with the prior art, the invention has the following beneficial effects: the method has the advantages that a large number of images can be labeled quickly, the image labeling speed and efficiency are improved, the label range is expanded, and the method is not limited to a trained target identification network.
Drawings
Fig. 1 is a schematic diagram of a picture to be labeled and a corresponding picture for automatic labeling according to an embodiment of the present invention.
FIG. 2 is a diagram illustrating an image for human inspection according to an embodiment of the present invention.
FIG. 3 is a diagram illustrating an automatically labeling csv file according to an embodiment of the present invention.
Detailed Description
The technical scheme of the invention is specifically explained below with reference to the accompanying drawings.
The invention provides an automatic image annotation system for new retail, which comprises:
the device comprises a test unit, a platform, a camera, a target and a test unit, wherein the test unit is used for obtaining a picture to be marked and the picture to be marked, which corresponds to the picture to be marked and is used for automatically marking the picture;
the cascade network testing unit is used for acquiring the position information of a target in a picture to be marked and a label of the target, acquiring the position information of the target in the picture through a trained image segmentation network aiming at the target, identifying numbers placed on the target through a deep learning network, and acquiring the label of the target through the corresponding relation between the numbers and the label;
the test unit is used for checking whether the marking result is a group Truth result or not, and is used for confirming whether the marking result is accurate or not, checking whether the position information has errors or not when the position information in the picture to be marked is acquired by using an image segmentation network, selecting the wrong picture for marking if the position information has errors, displaying each target and the number corresponding to each target after the number on the target is identified, detecting whether the number has errors or not, and modifying the label of the target if the number has errors;
and the test unit is used for generating a csv file, and writing the position information of each target of each picture, the label of the target and the name of the picture into the csv format file through the written python code and storing the csv format file.
The invention also provides a new retail image automatic labeling method, which comprises the following steps:
step S1: establishing a data set for a target to be marked, and training an image segmentation model through the data set to acquire position information of the target to be marked;
step S2: through a device with a fixed relative position between a platform and a camera, after a picture to be marked is shot, a paper sheet with numbers is placed on a target, and the picture to be marked is shot again to obtain a corresponding picture for automatic marking of the picture to be marked, so that a data set to be marked is manufactured;
step S3: obtaining the position information of each target in the corresponding picture for automatic annotation of the annotated picture through the image segmentation model in the step S1, checking whether the position information is wrong, and if the position information is wrong, continuing to annotate the annotated picture and saving the annotated picture in a folder for saving the picture to be annotated;
step S4: intercepting the image of each target in the image according to the position information of each target, identifying the number on the target through a deep learning network, and confirming the label of the target through the corresponding relation between the number and the label after checking that no error exists;
step S5: and compiling python codes, and making the file name of each picture, the position information and the label of each target in the picture into a csv format file to finish automatic labeling.
The following is a specific implementation of the present invention.
The invention relates to an automatic image labeling system for new retail, which comprises four parts, wherein the first part is a test unit for acquiring a picture to be labeled and the picture to be labeled, which corresponds to the picture to be labeled and is used for automatically labeling the picture, the second part is a test unit for acquiring position information of a target and a cascade network of labels of the target in the picture to be labeled, the third part is a test unit for checking whether a labeling result is a group Truth, and the fourth part is a test unit for acquiring the labeling result. The method comprises the steps of obtaining a picture to be marked and a test unit which corresponds to the picture to be marked and is used for automatically marking the picture, wherein the test unit mainly comprises the steps of shooting the picture to be marked, the picture which corresponds to the picture to be marked and is used for automatically marking, and establishing the corresponding relation between the picture to be marked and the picture to be marked. The test unit for acquiring the position information of the target and the label of the target in the picture to be marked mainly comprises the steps of acquiring the position information of the target in the picture, identifying the number on the target and acquiring the label of the target. The test unit for checking whether the labeling result is the group Truth includes checking whether the position information and the label of the automatically labeled target are wrong. The test unit for obtaining the labeling result comprises the steps of writing the position information of each target of each picture, the label of the target and the name of the picture into a csv format file and storing (as shown in fig. 3).
As shown in fig. 1, for the above system, the invention provides an automatic image annotation method for new retail, which is implemented according to the following steps:
step S1: the method comprises the steps of taking pictures to establish a data set aiming at a target needing to be trained, training an open-source image segmentation model (such as ICNET) by using the data set, and using the trained model to obtain position information of the target in a picture subsequently.
Step S2: through a device with a fixed relative position of a platform and a camera, after a picture to be marked is shot, a paper sheet with numbers is placed on a target, the picture to be marked is shot again to obtain a corresponding picture for automatic marking of the picture to be marked, each target category corresponds to a label with numbers for confirming the target, a data set to be marked is manufactured, and the corresponding relation between the two data sets is established
Step S3: acquiring the position information of each target in the corresponding picture for automatic annotation of the annotated picture through the image segmentation model (ICNET) in the step S1, manually checking whether the annotated picture has errors, and if the annotated picture has errors, not continuing to annotate the annotated picture and saving the annotated picture in a folder for storing the picture needing manual annotation;
step S4: intercepting images of all targets in the images, identifying numbers on the targets through a deep learning network, confirming the labels of the targets through the corresponding relation between the numbers and the labels after artificially checking the targets without errors, and manually modifying the labels of the targets if the targets are wrong.
Further, in step S2, the method further includes the following steps:
step S201: two folders are established for each picture to be marked, one folder stores the picture to be marked, the other folder stores the picture of a paper sheet with a figure on a target, the two pictures are shot by equipment with fixed relative positions of a camera and a platform for placing the target, and the names of the two pictures are the same.
Further, in step S3, the method further includes the following steps:
step S301: by positional information of the respective targets: (
,
,
,
) Adding a rectangular frame on each target on the picture, displaying the framed picture, storing the picture to be detected into a folder for storing the picture to be manually marked when the position information of the target is found to be wrongly marked or the position information of all the targets is not obtained by manual inspection, and carrying out the next step if the picture is correct.
Further, in step S4, the method further includes the following steps:
step S401: by positional information of the respective targets: (
,
,
,
) And intercepting pictures of each target in the targets, identifying the numbers of the cards with the numbers on the targets through a trained deep learning network (LENET-5), and labeling the targets according to the corresponding relation between the numbers and the labels.
Step S402: during the artificial inspection, the image of each target intercepted in the step S401 in the image is displayed, the recognized number is displayed at the upper left corner of the image of each target, the image is sent to the next step after the artificial inspection is correct, and if the image has errors, the label of the target is manually modified.
FIG. 2 is a diagram illustrating an image for human inspection according to an embodiment of the present invention.
The above are preferred embodiments of the present invention, and all changes made according to the technical scheme of the present invention that produce functional effects do not exceed the scope of the technical scheme of the present invention belong to the protection scope of the present invention.
Claims (5)
1. An automatic image annotation system for new retail sales, comprising:
the device comprises a test unit, a platform, a camera, a target and a test unit, wherein the test unit is used for obtaining a picture to be marked and the picture to be marked, which corresponds to the picture to be marked and is used for automatically marking the picture;
the cascade network testing unit is used for acquiring the position information of a target in a picture to be marked and a label of the target, acquiring the position information of the target in the picture through a trained image segmentation network aiming at the target, identifying numbers placed on the target through a deep learning network, and acquiring the label of the target through the corresponding relation between the numbers and the label;
the test unit is used for checking whether the marking result is a group Truth result or not, and is used for confirming whether the marking result is accurate or not, checking whether the position information has errors or not when the position information in the picture to be marked is acquired by using an image segmentation network, selecting the wrong picture for marking if the position information has errors, displaying each target and the number corresponding to each target after the number on the target is identified, detecting whether the number has errors or not, and modifying the label of the target if the number has errors;
and the test unit is used for generating a csv file, and writing the position information of each target of each picture, the label of the target and the name of the picture into the csv format file through the written python code and storing the csv format file.
2. An automatic image labeling method for new retail sale is characterized by comprising the following steps:
step S1: establishing a data set for a target to be marked, and training an image segmentation model through the data set to acquire position information of the target to be marked;
step S2: through a device with a fixed relative position between a platform and a camera, after a picture to be marked is shot, a paper sheet with numbers is placed on a target, and the picture to be marked is shot again to obtain a corresponding picture for automatic marking of the picture to be marked, so that a data set to be marked is manufactured;
step S3: obtaining the position information of each target in the corresponding picture for automatic annotation of the annotated picture through the image segmentation model in the step S1, checking whether the position information is wrong, and if the position information is wrong, continuing to annotate the annotated picture and saving the annotated picture in a folder for saving the picture to be annotated;
step S4: intercepting the image of each target in the image according to the position information of each target, identifying the number on the target through a deep learning network, and confirming the label of the target through the corresponding relation between the number and the label after checking that no error exists;
step S5: and compiling python codes, and making the file name of each picture, the position information and the label of each target in the picture into a csv format file to finish automatic labeling.
3. The automatic labeling method for image facing new retail sale of claim 2, wherein in the step S2, the method further comprises the following steps:
step S201: two folders are established for each picture to be marked, one folder stores the picture to be marked, the other folder stores the picture of the paper sheet with the number written on the target, the two pictures are shot by equipment with fixed relative positions of a camera and a platform for placing the target, and the names of the two pictures are the same.
4. The automatic labeling method for image facing new retail sale of claim 2, wherein in the step S3, the method further comprises the following steps:
step S301: and adding a rectangular frame to each target according to the position information of each target, displaying the framed picture, storing the picture to be detected into a folder for storing the picture to be marked if the picture is checked to be correct, and performing the next step if the picture is correct.
5. The automatic labeling method for image facing new retail sale of claim 2, wherein in the step S4, the method further comprises the following steps:
step S401: intercepting pictures of each target in the targets according to the position information of each target, identifying the numbers of the cards with the numbers written on the targets through a trained deep learning network, and labeling the targets in the area according to the corresponding relation between the numbers and the labels;
step S402: during checking, the picture of each target intercepted in the step S401 in the picture is displayed, the recognized number is displayed at the upper left corner of the picture of each target, and the next step is carried out after checking is correct.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010122977.6A CN111259184B (en) | 2020-02-27 | 2020-02-27 | Image automatic labeling system and method for new retail |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010122977.6A CN111259184B (en) | 2020-02-27 | 2020-02-27 | Image automatic labeling system and method for new retail |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111259184A true CN111259184A (en) | 2020-06-09 |
| CN111259184B CN111259184B (en) | 2022-03-08 |
Family
ID=70954571
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010122977.6A Active CN111259184B (en) | 2020-02-27 | 2020-02-27 | Image automatic labeling system and method for new retail |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111259184B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112734804A (en) * | 2021-01-07 | 2021-04-30 | 支付宝(杭州)信息技术有限公司 | System and method for image data annotation |
| CN113077421A (en) * | 2021-03-22 | 2021-07-06 | 南京南瑞继保电气有限公司 | Sample obtaining and rapid labeling method with relatively fixed target state |
| CN113095338A (en) * | 2021-06-10 | 2021-07-09 | 季华实验室 | Automatic labeling method and device for industrial product image, electronic equipment and storage medium |
Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060288842A1 (en) * | 1996-07-10 | 2006-12-28 | Sitrick David H | System and methodology for image and overlaid annotation display, management and communicaiton |
| CN103605984A (en) * | 2013-11-14 | 2014-02-26 | 厦门大学 | Supergraph learning-based indoor scene classification method |
| CN105389584A (en) * | 2015-10-13 | 2016-03-09 | 西北工业大学 | Streetscape semantic annotation method based on convolutional neural network and semantic transfer conjunctive model |
| CN108830154A (en) * | 2018-05-10 | 2018-11-16 | 明伟杰 | A kind of food nourishment composition detection method and system based on binocular camera |
| CN108985293A (en) * | 2018-06-22 | 2018-12-11 | 深源恒际科技有限公司 | A kind of image automation mask method and system based on deep learning |
| CN108985214A (en) * | 2018-07-09 | 2018-12-11 | 上海斐讯数据通信技术有限公司 | The mask method and device of image data |
| CN109255044A (en) * | 2018-08-31 | 2019-01-22 | 江苏大学 | A kind of image intelligent mask method based on YOLOv3 deep learning network |
| CN109284686A (en) * | 2018-08-23 | 2019-01-29 | 国网山西省电力公司计量中心 | A kind of label identification method that camera automatic pitching is taken pictures |
| CN109446369A (en) * | 2018-09-28 | 2019-03-08 | 武汉中海庭数据技术有限公司 | The exchange method and system of the semi-automatic mark of image |
| CN109543596A (en) * | 2018-11-20 | 2019-03-29 | 浙江大华技术股份有限公司 | A kind of water level monitoring method, apparatus, electronic equipment and storage medium |
| CN109725733A (en) * | 2019-01-25 | 2019-05-07 | 中国人民解放军国防科技大学 | Human-computer interaction method and human-computer interaction equipment based on augmented reality |
| CN110210487A (en) * | 2019-05-30 | 2019-09-06 | 上海商汤智能科技有限公司 | A kind of image partition method and device, electronic equipment and storage medium |
| CN110211134A (en) * | 2019-05-30 | 2019-09-06 | 上海商汤智能科技有限公司 | A kind of image partition method and device, electronic equipment and storage medium |
| CN110320921A (en) * | 2019-08-07 | 2019-10-11 | 厦门大学 | A kind of shared intelligent wheelchair system of the self-navigation towards hospital's traffic |
| CN110705335A (en) * | 2018-07-10 | 2020-01-17 | 北京京东尚科信息技术有限公司 | Method and device for labeling sample picture |
| US20200034693A1 (en) * | 2018-07-27 | 2020-01-30 | Samsung Electronics Co., Ltd. | Method for detecting defects in semiconductor device |
-
2020
- 2020-02-27 CN CN202010122977.6A patent/CN111259184B/en active Active
Patent Citations (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060288842A1 (en) * | 1996-07-10 | 2006-12-28 | Sitrick David H | System and methodology for image and overlaid annotation display, management and communicaiton |
| CN103605984A (en) * | 2013-11-14 | 2014-02-26 | 厦门大学 | Supergraph learning-based indoor scene classification method |
| CN105389584A (en) * | 2015-10-13 | 2016-03-09 | 西北工业大学 | Streetscape semantic annotation method based on convolutional neural network and semantic transfer conjunctive model |
| CN108830154A (en) * | 2018-05-10 | 2018-11-16 | 明伟杰 | A kind of food nourishment composition detection method and system based on binocular camera |
| CN108985293A (en) * | 2018-06-22 | 2018-12-11 | 深源恒际科技有限公司 | A kind of image automation mask method and system based on deep learning |
| CN108985214A (en) * | 2018-07-09 | 2018-12-11 | 上海斐讯数据通信技术有限公司 | The mask method and device of image data |
| CN110705335A (en) * | 2018-07-10 | 2020-01-17 | 北京京东尚科信息技术有限公司 | Method and device for labeling sample picture |
| US20200034693A1 (en) * | 2018-07-27 | 2020-01-30 | Samsung Electronics Co., Ltd. | Method for detecting defects in semiconductor device |
| CN109284686A (en) * | 2018-08-23 | 2019-01-29 | 国网山西省电力公司计量中心 | A kind of label identification method that camera automatic pitching is taken pictures |
| CN109255044A (en) * | 2018-08-31 | 2019-01-22 | 江苏大学 | A kind of image intelligent mask method based on YOLOv3 deep learning network |
| CN109446369A (en) * | 2018-09-28 | 2019-03-08 | 武汉中海庭数据技术有限公司 | The exchange method and system of the semi-automatic mark of image |
| CN109543596A (en) * | 2018-11-20 | 2019-03-29 | 浙江大华技术股份有限公司 | A kind of water level monitoring method, apparatus, electronic equipment and storage medium |
| CN109725733A (en) * | 2019-01-25 | 2019-05-07 | 中国人民解放军国防科技大学 | Human-computer interaction method and human-computer interaction equipment based on augmented reality |
| CN110210487A (en) * | 2019-05-30 | 2019-09-06 | 上海商汤智能科技有限公司 | A kind of image partition method and device, electronic equipment and storage medium |
| CN110211134A (en) * | 2019-05-30 | 2019-09-06 | 上海商汤智能科技有限公司 | A kind of image partition method and device, electronic equipment and storage medium |
| CN110320921A (en) * | 2019-08-07 | 2019-10-11 | 厦门大学 | A kind of shared intelligent wheelchair system of the self-navigation towards hospital's traffic |
Non-Patent Citations (2)
| Title |
|---|
| M GUILLAUMIN ET AL.: ""Imagenet auto-annotation with segmentation propagation"", 《SPRINGER》 * |
| 柯逍 等: ""基于Co-training的图像自动标注"", 《厦门大学学报(自然科学版)》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112734804A (en) * | 2021-01-07 | 2021-04-30 | 支付宝(杭州)信息技术有限公司 | System and method for image data annotation |
| CN112734804B (en) * | 2021-01-07 | 2022-08-26 | 支付宝(杭州)信息技术有限公司 | System and method for image data annotation |
| CN113077421A (en) * | 2021-03-22 | 2021-07-06 | 南京南瑞继保电气有限公司 | Sample obtaining and rapid labeling method with relatively fixed target state |
| CN113095338A (en) * | 2021-06-10 | 2021-07-09 | 季华实验室 | Automatic labeling method and device for industrial product image, electronic equipment and storage medium |
| CN113095338B (en) * | 2021-06-10 | 2021-11-09 | 季华实验室 | Automatic labeling method and device for industrial product image, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111259184B (en) | 2022-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111259184B (en) | Image automatic labeling system and method for new retail | |
| CN110659662B (en) | Flaw detection system and method using artificial intelligence | |
| CN108985199A (en) | Detection method, device and the storage medium of commodity loading or unloading operation | |
| WO2017092427A1 (en) | Electronic element positioning method and apparatus | |
| CN105184778A (en) | Detection method and apparatus | |
| WO2016104965A1 (en) | Learning assessment service system | |
| CN107907775B (en) | Wiring detection method and wiring detection device | |
| CN109359043B (en) | Mobile game automatic testing method based on machine learning | |
| KR102195999B1 (en) | Method, device and system for processing image tagging information | |
| CN115239683A (en) | Detection method of circuit board, model training method and device and electronic equipment | |
| CN110728269A (en) | High-speed rail contact net support pole number plate identification method | |
| CN112182266A (en) | Picture labeling and classifying method based on picture numbers | |
| CN116128853A (en) | Production line assembly detection method, system, computer and readable storage medium | |
| TWI823463B (en) | Label integrity adaptive detection method and system | |
| CN115908977A (en) | Image data labeling method and device, electronic equipment and storage medium | |
| CN110956087A (en) | Method and device for identifying table in picture, readable medium and electronic equipment | |
| CN113807434B (en) | Cloth flaw identification method and model training method | |
| US11481577B2 (en) | Machine learning (ML) quality assurance for data curation | |
| CN115457585A (en) | Processing method and device for homework correction, computer equipment and readable storage medium | |
| JP4160206B2 (en) | Database registration method using character recognition device | |
| CN113516131A (en) | Image processing method, device, equipment and storage medium | |
| CN111674033A (en) | Engine nameplate intelligent printing and sorting method and system based on OCR | |
| CN111563407A (en) | Model training method, and image direction correcting method and device | |
| CN112529039B (en) | Method and device for checking material information of main board and storage medium | |
| WO2022060082A1 (en) | Device and method for collecting reagent information by means of image analysis using reagent bottle features |
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 |