CN117115162A - Chip production control system for detecting battery based on visual analysis - Google Patents
Chip production control system for detecting battery based on visual analysis Download PDFInfo
- Publication number
- CN117115162A CN117115162A CN202311376720.3A CN202311376720A CN117115162A CN 117115162 A CN117115162 A CN 117115162A CN 202311376720 A CN202311376720 A CN 202311376720A CN 117115162 A CN117115162 A CN 117115162A
- Authority
- CN
- China
- Prior art keywords
- battery
- module
- target
- yolov5
- battery surface
- 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
- 238000004519 manufacturing process Methods 0.000 title claims abstract description 64
- 238000004458 analytical method Methods 0.000 title claims abstract description 28
- 230000000007 visual effect Effects 0.000 title claims abstract description 27
- 238000001514 detection method Methods 0.000 claims abstract description 197
- 230000007547 defect Effects 0.000 claims abstract description 152
- 238000012549 training Methods 0.000 claims abstract description 30
- 239000012634 fragment Substances 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 11
- 238000003709 image segmentation Methods 0.000 claims description 11
- 238000012545 processing Methods 0.000 claims description 11
- 230000002950 deficient Effects 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000000137 annealing Methods 0.000 claims description 6
- 238000000034 method Methods 0.000 claims description 4
- 230000011218 segmentation Effects 0.000 claims description 4
- 210000004027 cell Anatomy 0.000 description 11
- 238000010586 diagram Methods 0.000 description 6
- 239000000463 material Substances 0.000 description 2
- 238000000053 physical method Methods 0.000 description 2
- 208000003464 asthenopia Diseases 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000009776 industrial production Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000011897 real-time detection Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/40—Image enhancement or restoration using histogram techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/70—Determining position or orientation of objects or cameras
- G06T7/73—Determining position or orientation of objects or cameras using feature-based methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/774—Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20081—Training; Learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20084—Artificial neural networks [ANN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30108—Industrial image inspection
- G06T2207/30148—Semiconductor; IC; Wafer
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- General Health & Medical Sciences (AREA)
- Multimedia (AREA)
- Computational Linguistics (AREA)
- Medical Informatics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Molecular Biology (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Quality & Reliability (AREA)
- Investigating Materials By The Use Of Optical Means Adapted For Particular Applications (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the field of image recognition, in particular to a chip production control system for detecting a battery based on visual analysis. The chip image module is used for obtaining image data of the surface of the battery to be trained; the model building module is used for obtaining an initial YOLOv5 battery surface detection model; the model training module is used for inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model; the battery detection module is used for obtaining the surface defect state of the target battery; the chip positioning module is used for acquiring the position information of the target battery through the chip positioning module if the surface defect state of the target battery is a serious defect state, and controlling the target battery through the chip control module; and the chip control module is used for controlling the overturning of the target battery through the chip control module. The accuracy and precision of the detection of the surface defects of the battery can be improved.
Description
Technical Field
The invention relates to the field of image recognition, in particular to a chip production control system for detecting a battery based on visual analysis.
Background
In modern production systems with increasingly popular automation, most factories still adopt an artificial visual detection method for quality inspection of batteries. Workers combine existing specifications and working experience to find some common and obvious defects with naked eyes. However, the manual visual detection method is still limited by subjective factors, visual fatigue and other problems, so that the quality evaluation efficiency of the battery plate is low and inaccurate, the real-time detection of the battery plate can not be realized, and the popularization and application of the battery plate are limited. At present, another detection method applied to industrial production is a manual physical method, but the type and the position of defects cannot be judged, and the detection speed is low, so that the actual production requirements cannot be well met. In addition to the artificial visual detection method and the artificial physical method, another main detection method is to perform image analysis by using machine vision. Because the accuracy of the machine vision detection method needs to be improved, the labor is saved, the productivity is improved, and the development of a rapid and precise solar panel defect detection technology is needed.
Disclosure of Invention
The invention aims to solve the problems, and designs a chip production control system for detecting a battery based on visual analysis.
The technical scheme of the invention for achieving the purpose is that in the chip production control system for detecting the battery based on visual analysis, the chip production control system comprises the following steps:
the chip image module is used for acquiring historical battery surface image data corresponding to a battery in the system, and carrying out data preprocessing on the historical battery surface image data to obtain battery surface image data to be trained;
the model building module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm, replacing a C3 structure of the YOLOv5 battery surface detection model by using a Ghost Block light module, and introducing a CBAM attention mechanism to obtain an initial YOLOv5 battery surface detection model;
the model training module is used for inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model, and the target Yolov5 battery surface detection model is arranged in the chip image module;
the battery detection module is used for acquiring real-time battery surface image data of a target battery in a production line, inputting the real-time battery surface image data into the target YOLOv5 battery surface detection model for identification, and obtaining a target battery surface defect state;
the chip positioning module is used for acquiring the position information of the target battery through the chip positioning module if the surface defect state of the target battery is a serious defect state, and controlling the target battery through the chip control module;
and the chip control module is used for performing overturn control on the target battery through the chip control module and performing surface defect detection on the target battery through the chip image module if the surface defect state of the target battery is a general defect state.
Further, in the above chip production control system, the chip image module includes an image acquisition unit, an image enhancement unit, a feature extraction unit, an image segmentation unit:
the image acquisition unit is used for acquiring historical battery surface image data corresponding to a battery in the system through the image acquisition device, wherein the historical battery surface image data at least comprises a battery hidden crack defect, a battery fragment defect, a battery broken gate defect and a battery black core defect, and the image acquisition device at least comprises a CCD image sensor and a CMOS image sensor;
the image enhancement unit is used for carrying out image enhancement on the historical battery surface image data by utilizing histogram equalization to obtain enhanced battery surface image data;
the feature extraction unit is used for carrying out feature extraction on the enhanced battery surface image data based on the local binary pattern to obtain feature battery surface image data;
and the image segmentation unit is used for carrying out image segmentation on the characteristic battery surface image data by utilizing a motion analysis segmentation algorithm to obtain battery surface image data to be trained.
Further, in the above chip production control system, the model building module includes a building sub-module, a replacing sub-module, a parameter sub-module, an introducing sub-module, a processing sub-module, and an obtaining sub-module:
the building sub-module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm;
the replacing sub-module is used for replacing the C3 structure of the Yolov5 battery surface detection model by using a Ghost Block light module;
a parameter submodule for reducing parameters of the YOLOv5 battery surface detection model by decomposing standard convolution to generate an inexpensive feature map;
the introducing submodule is used for introducing a group of detection anchor frames on the low-layer feature map of the Yolov5 battery surface detection model;
the processing sub-module is used for independently processing the Yolov5 battery surface detection model by utilizing the decoupling detection head to carry out regression and classification of the anchor frame;
and the obtained submodule is used for introducing a CBAM attention mechanism to the Yolov5 battery surface detection model to obtain an initial Yolov5 battery surface detection model.
Further, in the above chip production control system, the model training module is characterized in that it includes a model training unit, a learning rate adjusting unit, a threshold confidence unit, a model setting unit:
the model training unit is used for inputting the image data of the battery surface to be trained into the initial YOLOv5 battery surface detection model for training;
the learning rate adjusting unit is used for adjusting the learning rate of the initial YOLOv5 battery surface detection model by using a cosine annealing algorithm, so that the initial learning rate=0.01 and the final learning rate=0.1;
a threshold confidence unit, configured to set a threshold confidence=0.25 and an nms threshold=0.45 of the initial YOLOv5 battery surface detection model, so as to obtain a target YOLOv5 battery surface detection model;
and the model setting unit is used for setting the target YOLOv5 battery surface detection model in the chip image module and detecting battery surface image data.
Further, in the above chip production control system, the battery detection module includes a detection sub-module, a state determination sub-module, a serious defect module, a general defect module, and a defect-free module:
the detection sub-module is used for acquiring real-time battery surface image data of a target battery in a production line, inputting the real-time battery surface image data into the target YOLOv5 battery surface detection model for identification, and obtaining a target battery surface defect state;
a state determination sub-module for determining that the target battery surface defect state includes at least a serious defect state, a general defect state, and a non-defective state;
a serious defect module for determining that the serious defect state comprises a target battery including a battery hidden crack defect, a battery fragment defect, a battery broken gate defect and a battery black core defect;
a general defect module for determining that the general defect state includes a target battery including one of a battery hidden crack defect, a battery fragment defect, a battery gate-break defect, and a battery black core defect;
a defect-free module for determining that the defect-free state includes that the target cell does not contain a cell hidden crack defect, a cell fragment defect, a cell gate break defect, and a cell black core defect.
Further, in the above chip production control system, the chip positioning module includes a position positioning unit, an instruction generating unit, a production control unit:
the device comprises a position locating unit, a chip locating module and a control unit, wherein the position locating unit is used for obtaining the position information of the target battery through the chip locating module if the surface defect state of the target battery is a serious defect state, and the chip locating module at least comprises a CCD image sensor and a CMOS image sensor;
the instruction generation unit is used for sending the position information of the target battery to the chip control module to generate a target battery control instruction;
and the production control unit is used for controlling the target battery through a target battery control instruction, wherein the target battery control instruction at least comprises the steps of removing the target battery from the production line and transferring the target battery from the production line to a secondary detection warehouse.
Further, in the above chip production control system, the chip control module includes a control sub-module, a detection sub-module, a removal sub-module:
the control sub-module is used for performing overturn control on the target battery through the chip control module if the surface defect state of the target battery is a general defect state;
the detection sub-module is used for acquiring second real-time battery surface image data of the target battery after being overturned by using the chip image module, and inputting the second real-time battery surface image data into a target YOLOv5 battery surface detection model for detection to obtain a second target battery surface defect state;
and the removing sub-module is used for removing the target battery through the chip control module if the surface defect state of the second target battery is a general defect state.
Further, in the above chip production control system, the chip production control system further includes the steps of:
acquiring historical battery surface image data corresponding to a battery in a system, and performing data preprocessing on the historical battery surface image data to obtain battery surface image data to be trained;
a YOLOv5 battery surface detection model is established through a YOLO target detection algorithm, a Ghost Block light module is utilized to replace a C3 structure of the YOLOv5 battery surface detection model, a CBAM attention mechanism is introduced, and an initial YOLOv5 battery surface detection model is obtained;
inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model, wherein the target Yolov5 battery surface detection model is arranged in a chip image module;
acquiring real-time battery surface image data of a target battery in a production line, and inputting the real-time battery surface image data into a target YOLOv5 battery surface detection model for identification to obtain a target battery surface defect state;
if the surface defect state of the target battery is a serious defect state, acquiring the position information of the target battery through a chip positioning module, and controlling the target battery through a chip control module;
if the surface defect state of the target battery is a general defect state, the target battery is subjected to overturn control by a chip control module, and the chip image module is used for detecting the surface defect of the target battery.
Further, in the above chip production control system, the chip production control system further includes the steps of:
building a YOLOv5 battery surface detection model through a YOLO target detection algorithm;
replacing a C3 structure of the YOLOv5 battery surface detection model by using a Ghost Block light module;
reducing parameters of the YOLOv5 battery surface detection model by decomposing standard convolution to generate an inexpensive feature map;
introducing a group of detection anchor frames on a low-level feature map of the Yolov5 battery surface detection model;
the decoupling detection head is utilized to independently process the regression and classification of the anchor frame to the Yolov5 battery surface detection model;
introducing a CBAM attention mechanism to the YOLOv5 battery surface detection model to obtain an initial YOLOv5 battery surface detection model.
Further, in the above chip production control system, the chip production control system further includes the steps of:
inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training;
adjusting the learning rate of the initial YOLOv5 battery surface detection model by using a cosine annealing algorithm, wherein the initial learning rate=0.01 and the final learning rate=0.1;
setting the threshold confidence coefficient=0.25 and NMS threshold value=0.45 of the initial Yolov5 battery surface detection model to obtain a target Yolov5 battery surface detection model;
and setting the target YOLOv5 battery surface detection model in a chip image module for detecting battery surface image data.
The method has the advantages that the chip image module is used for acquiring historical battery surface image data corresponding to the battery in the system, and carrying out data preprocessing on the historical battery surface image data to obtain battery surface image data to be trained; the model building module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm, replacing a C3 structure of the YOLOv5 battery surface detection model by using a Ghost Block light module, and introducing a CBAM attention mechanism to obtain an initial YOLOv5 battery surface detection model; the model training module is used for inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model, and the target Yolov5 battery surface detection model is arranged in the chip image module; the battery detection module is used for acquiring real-time battery surface image data of a target battery in a production line, inputting the real-time battery surface image data into the target YOLOv5 battery surface detection model for identification, and obtaining a target battery surface defect state; the chip positioning module is used for acquiring the position information of the target battery through the chip positioning module if the surface defect state of the target battery is a serious defect state, and controlling the target battery through the chip control module; and the chip control module is used for performing overturn control on the target battery through the chip control module and performing surface defect detection on the target battery through the chip image module if the surface defect state of the target battery is a general defect state. The accuracy and precision of battery surface defect detection can be improved, defective batteries in a production line can be positioned in real time, the defective batteries can be adjusted and removed in time, the battery detection effectiveness is guaranteed, and the use of manpower and material resources is reduced.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the invention.
FIG. 1 is a schematic diagram of a first embodiment of a chip production control system for detecting a battery based on visual analysis in an embodiment of the present invention;
FIG. 2 is a schematic diagram of a second embodiment of a chip production control system for detecting a battery based on visual analysis in an embodiment of the present invention;
FIG. 3 is a schematic diagram of a third embodiment of a chip production control system for detecting a battery based on visual analysis in an embodiment of the present invention;
fig. 4 is a schematic diagram of a fourth embodiment of a chip production control system for detecting a battery based on visual analysis in an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the drawings and examples, in order to make the objects, technical solutions and advantages of the present invention more apparent. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
As used herein, the singular forms "a", "an", "the" and "the" are intended to include the plural forms as well, unless expressly stated otherwise, as understood by those skilled in the art. It will be further understood that the terms "comprises" and/or "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
The present invention will be described in detail with reference to the accompanying drawings, as shown in fig. 1, a chip production control system for detecting a battery based on visual analysis, the chip production control system comprising the steps of:
the chip image module is used for acquiring historical battery surface image data corresponding to a battery in the system, and carrying out data preprocessing on the historical battery surface image data to obtain battery surface image data to be trained;
specifically, the embodiment includes an image acquisition unit, configured to acquire, by using an image acquisition device, historical battery surface image data corresponding to a battery in a system, where the historical battery surface image data includes at least a battery hidden crack defect, a battery fragment defect, a battery gate breakage defect, and a battery black core defect, and the image acquisition device includes at least a CCD image sensor and a CMOS image sensor; the image enhancement unit is used for carrying out image enhancement on the historical battery surface image data by utilizing histogram equalization to obtain enhanced battery surface image data; the feature extraction unit is used for carrying out feature extraction on the enhanced battery surface image data based on the local binary pattern to obtain feature battery surface image data; and the image segmentation unit is used for carrying out image segmentation on the characteristic battery surface image data by utilizing a motion analysis segmentation algorithm to obtain battery surface image data to be trained.
The model building module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm, replacing a C3 structure of the YOLOv5 battery surface detection model by using a Ghost Block light module, and introducing a CBAM attention mechanism to obtain an initial YOLOv5 battery surface detection model;
specifically, the embodiment includes a building sub-module for building a YOLOv5 battery surface detection model through a YOLOv target detection algorithm; the replacement sub-module is used for replacing a C3 structure of the YOLOv5 battery surface detection model by using the Ghost Block light module; the parameter submodule is used for reducing the parameters of the YOLOv5 battery surface detection model in a mode of decomposing the standard convolution to generate an inexpensive characteristic diagram; the introduction sub-module is used for introducing a group of detection anchor frames on a low-layer feature map of the YOLOv5 battery surface detection model; the processing submodule is used for independently processing the detection model of the surface of the YOLOv5 battery by utilizing the decoupling detection head to carry out regression and classification of the anchor frame; and obtaining a sub-module for introducing a CBAM attention mechanism to the YOLOv5 battery surface detection model to obtain an initial YOLOv5 battery surface detection model.
The model training module is used for inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model, and the target Yolov5 battery surface detection model is arranged in the chip image module;
specifically, the embodiment includes a model training unit, configured to input image data of a battery surface to be trained into an initial YOLOv5 battery surface detection model for training; the learning rate adjusting unit is used for adjusting the learning rate of the initial YOLOv5 battery surface detection model by using a cosine annealing algorithm, so that the initial learning rate=0.01 and the final learning rate=0.1; a threshold confidence unit, configured to set a threshold confidence=0.25 and an nms threshold=0.45 of the initial YOLOv5 battery surface detection model, so as to obtain a target YOLOv5 battery surface detection model; and the model setting unit is used for setting the target YOLOv5 battery surface detection model in the chip image module and detecting battery surface image data.
The battery detection module is used for acquiring real-time battery surface image data of a target battery in the production line, inputting the real-time battery surface image data into a target YOLOv5 battery surface detection model for identification, and obtaining a target battery surface defect state;
specifically, the embodiment includes a detection sub-module, configured to obtain real-time battery surface image data of a target battery in a production line, input the real-time battery surface image data to a target YOLOv5 battery surface detection model for identification, and obtain a target battery surface defect state; a state determination sub-module for determining that the target cell surface defect state includes at least a serious defect state, a general defect state, and a non-defective state; a serious defect module for determining a serious defect state including a target battery including a battery hidden crack defect, a battery fragment defect, a battery gate breakage defect, and a battery black core defect; a general defect module for determining that the general defect state includes a target battery including one of a battery hidden crack defect, a battery fragment defect, a battery gate-broken defect, and a battery black core defect; a defect-free module for determining defect-free conditions including a target cell not including a cell hidden crack defect, a cell fragment defect, a cell gate break defect, and a cell black core defect.
The chip positioning module is used for acquiring the position information of the target battery through the chip positioning module if the surface defect state of the target battery is a serious defect state, and controlling the target battery through the chip control module;
specifically, the embodiment includes a position locating unit, configured to obtain, if a surface defect state of a target battery of the target battery is a serious defect state, position information of the target battery through a chip locating module, where the chip locating module includes at least a CCD image sensor and a CMOS image sensor; the instruction generation unit is used for sending the position information of the target battery to the chip control module and generating a target battery control instruction; and the production control unit is used for controlling the target battery through a target battery control instruction, wherein the target battery control instruction at least comprises the steps of removing the target battery from the production line and transferring the target battery from the production line to a secondary detection warehouse.
And the chip control module is used for performing overturn control on the target battery through the chip control module and performing surface defect detection on the target battery through the chip image module if the surface defect state of the target battery is a general defect state.
Specifically, the embodiment includes a control submodule, configured to perform overturn control on the target battery through the chip control module if the surface defect state of the target battery is a general defect state; the detection sub-module is used for acquiring second real-time battery surface image data of the target battery after being overturned by using the chip image module, and inputting the second real-time battery surface image data into the target YOLOv5 battery surface detection model for detection to obtain a second target battery surface defect state; and the removing sub-module is used for removing the target battery through the chip control module if the surface defect state of the second target battery is a general defect state.
The method has the advantages that the chip image module is used for acquiring historical battery surface image data corresponding to the battery in the system, and carrying out data preprocessing on the historical battery surface image data to obtain battery surface image data to be trained; the model building module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm, replacing a C3 structure of the YOLOv5 battery surface detection model by using a Ghost Block light module, and introducing a CBAM attention mechanism to obtain an initial YOLOv5 battery surface detection model; the model training module is used for inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model, and the target Yolov5 battery surface detection model is arranged in the chip image module; the battery detection module is used for acquiring real-time battery surface image data of a target battery in the production line, inputting the real-time battery surface image data into a target YOLOv5 battery surface detection model for identification, and obtaining a target battery surface defect state; the chip positioning module is used for acquiring the position information of the target battery through the chip positioning module if the surface defect state of the target battery is a serious defect state, and controlling the target battery through the chip control module; and the chip control module is used for performing overturn control on the target battery through the chip control module and performing surface defect detection on the target battery through the chip image module if the surface defect state of the target battery is a general defect state. The accuracy and precision of battery surface defect detection can be improved, defective batteries in a production line can be positioned in real time, the defective batteries can be adjusted and removed in time, the battery detection effectiveness is guaranteed, and the use of manpower and material resources is reduced.
In this embodiment, referring to fig. 2, in a second embodiment of a chip production control system for detecting a battery based on visual analysis according to an embodiment of the present invention, a chip image module includes an image acquisition unit, an image enhancement unit, a feature extraction unit, and an image segmentation unit:
the image acquisition unit is used for acquiring historical battery surface image data corresponding to a battery in the system through the image acquisition device, wherein the historical battery surface image data at least comprises a battery hidden crack defect, a battery fragment defect, a battery broken gate defect and a battery black core defect, and the image acquisition device at least comprises a CCD image sensor and a CMOS image sensor;
the image enhancement unit is used for carrying out image enhancement on the historical battery surface image data by utilizing histogram equalization to obtain enhanced battery surface image data;
the feature extraction unit is used for carrying out feature extraction on the enhanced battery surface image data based on the local binary pattern to obtain feature battery surface image data;
and the image segmentation unit is used for carrying out image segmentation on the characteristic battery surface image data by utilizing a motion analysis segmentation algorithm to obtain battery surface image data to be trained.
In this embodiment, referring to fig. 3, in a third embodiment of a chip production control system for detecting a battery based on visual analysis in the embodiment of the present invention, a model building module includes a building sub-module, a replacing sub-module, a parameter sub-module, an introducing sub-module, a processing sub-module, and an obtaining sub-module:
the building sub-module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm;
the replacement sub-module is used for replacing a C3 structure of the YOLOv5 battery surface detection model by using the Ghost Block light module;
the parameter submodule is used for reducing the parameters of the YOLOv5 battery surface detection model in a mode of decomposing the standard convolution to generate an inexpensive characteristic diagram;
the introduction sub-module is used for introducing a group of detection anchor frames on a low-layer feature map of the YOLOv5 battery surface detection model;
the processing submodule is used for independently processing the detection model of the surface of the YOLOv5 battery by utilizing the decoupling detection head to carry out regression and classification of the anchor frame;
and obtaining a sub-module for introducing a CBAM attention mechanism to the YOLOv5 battery surface detection model to obtain an initial YOLOv5 battery surface detection model.
In this embodiment, referring to fig. 4, in a third embodiment of a chip production control system for detecting a battery based on visual analysis in the embodiment of the present invention, a model training module includes a model training unit, a learning rate adjusting unit, a threshold confidence unit, and a model setting unit:
the model training unit is used for inputting the image data of the battery surface to be trained into an initial Yolov5 battery surface detection model for training;
the learning rate adjusting unit is used for adjusting the learning rate of the initial YOLOv5 battery surface detection model by using a cosine annealing algorithm, so that the initial learning rate=0.01 and the final learning rate=0.1;
a threshold confidence unit, configured to set a threshold confidence=0.25 and an nms threshold=0.45 of the initial YOLOv5 battery surface detection model, so as to obtain a target YOLOv5 battery surface detection model;
and the model setting unit is used for setting the target YOLOv5 battery surface detection model in the chip image module and detecting battery surface image data.
The foregoing has shown and described the basic principles, principal features and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the above-described embodiments, and that the above-described embodiments and descriptions are only preferred embodiments of the present invention, and are not intended to limit the invention, and that various changes and modifications may be made therein without departing from the spirit and scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (10)
1. Chip production control system that detects the battery based on visual analysis, its characterized in that, chip production control system includes following module:
the chip image module is used for acquiring historical battery surface image data corresponding to a battery in the system, and carrying out data preprocessing on the historical battery surface image data to obtain battery surface image data to be trained;
the model building module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm, replacing a C3 structure of the YOLOv5 battery surface detection model by using a Ghost Block light module, and introducing a CBAM attention mechanism to obtain an initial YOLOv5 battery surface detection model;
the model training module is used for inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model, and the target Yolov5 battery surface detection model is arranged in the chip image module;
the battery detection module is used for acquiring real-time battery surface image data of a target battery in a production line, inputting the real-time battery surface image data into the target YOLOv5 battery surface detection model for identification, and obtaining a target battery surface defect state;
the chip positioning module is used for acquiring the position information of the target battery through the chip positioning module if the surface defect state of the target battery is a serious defect state, and controlling the target battery through the chip control module;
and the chip control module is used for performing overturn control on the target battery through the chip control module and performing surface defect detection on the target battery through the chip image module if the surface defect state of the target battery is a general defect state.
2. The chip production control system for detecting a battery based on visual analysis according to claim 1, wherein the chip image module comprises an image acquisition unit, an image enhancement unit, a feature extraction unit, an image segmentation unit:
the image acquisition unit is used for acquiring historical battery surface image data corresponding to a battery in the system through the image acquisition device, wherein the historical battery surface image data at least comprises a battery hidden crack defect, a battery fragment defect, a battery broken gate defect and a battery black core defect, and the image acquisition device at least comprises a CCD image sensor and a CMOS image sensor;
the image enhancement unit is used for carrying out image enhancement on the historical battery surface image data by utilizing histogram equalization to obtain enhanced battery surface image data;
the feature extraction unit is used for carrying out feature extraction on the enhanced battery surface image data based on the local binary pattern to obtain feature battery surface image data;
and the image segmentation unit is used for carrying out image segmentation on the characteristic battery surface image data by utilizing a motion analysis segmentation algorithm to obtain battery surface image data to be trained.
3. The chip production control system for detecting a battery based on visual analysis according to claim 1, wherein the model building module comprises a building sub-module, a replacing sub-module, a parameter sub-module, an introducing sub-module, a processing sub-module, and a obtaining sub-module:
the building sub-module is used for building a YOLOv5 battery surface detection model through a YOLO target detection algorithm;
the replacing sub-module is used for replacing the C3 structure of the Yolov5 battery surface detection model by using a Ghost Block light module;
a parameter submodule for reducing parameters of the YOLOv5 battery surface detection model by decomposing standard convolution to generate an inexpensive feature map;
the introducing submodule is used for introducing a group of detection anchor frames on the low-layer feature map of the Yolov5 battery surface detection model;
the processing sub-module is used for independently processing the Yolov5 battery surface detection model by utilizing the decoupling detection head to carry out regression and classification of the anchor frame;
and the obtained submodule is used for introducing a CBAM attention mechanism to the Yolov5 battery surface detection model to obtain an initial Yolov5 battery surface detection model.
4. The chip production control system for detecting a battery based on visual analysis according to claim 1, wherein the model training module comprises a model training unit, a learning rate adjusting unit, a threshold confidence unit, a model setting unit:
the model training unit is used for inputting the image data of the battery surface to be trained into the initial YOLOv5 battery surface detection model for training;
the learning rate adjusting unit is used for adjusting the learning rate of the initial YOLOv5 battery surface detection model by using a cosine annealing algorithm, so that the initial learning rate=0.01 and the final learning rate=0.1;
a threshold confidence unit, configured to set a threshold confidence=0.25 and an nms threshold=0.45 of the initial YOLOv5 battery surface detection model, so as to obtain a target YOLOv5 battery surface detection model;
and the model setting unit is used for setting the target YOLOv5 battery surface detection model in the chip image module and detecting battery surface image data.
5. The chip production control system for detecting a battery based on visual analysis according to claim 1, wherein the battery detection module comprises a detection sub-module, a state determination sub-module, a serious defect module, a general defect module, a no-defect module:
the detection sub-module is used for acquiring real-time battery surface image data of a target battery in a production line, inputting the real-time battery surface image data into the target YOLOv5 battery surface detection model for identification, and obtaining a target battery surface defect state;
a state determination sub-module for determining that the target battery surface defect state includes at least a serious defect state, a general defect state, and a non-defective state;
a serious defect module for determining that the serious defect state comprises a target battery including a battery hidden crack defect, a battery fragment defect, a battery broken gate defect and a battery black core defect;
a general defect module for determining that the general defect state includes a target battery including one of a battery hidden crack defect, a battery fragment defect, a battery gate-break defect, and a battery black core defect;
a defect-free module for determining that the defect-free state includes that the target cell does not contain a cell hidden crack defect, a cell fragment defect, a cell gate break defect, and a cell black core defect.
6. The chip production control system for detecting a battery based on visual analysis according to claim 1, wherein the chip positioning module comprises a position positioning unit, an instruction generating unit, a production control unit:
the device comprises a position locating unit, a chip locating module and a control unit, wherein the position locating unit is used for obtaining the position information of the target battery through the chip locating module if the surface defect state of the target battery is a serious defect state, and the chip locating module at least comprises a CCD image sensor and a CMOS image sensor;
the instruction generation unit is used for sending the position information of the target battery to the chip control module to generate a target battery control instruction;
and the production control unit is used for controlling the target battery through a target battery control instruction, wherein the target battery control instruction at least comprises the steps of removing the target battery from the production line and transferring the target battery from the production line to a secondary detection warehouse.
7. The chip production control system for detecting a battery based on visual analysis according to claim 1, wherein the chip control module comprises a control sub-module, a detection sub-module, a removal sub-module:
the control sub-module is used for performing overturn control on the target battery through the chip control module if the surface defect state of the target battery is a general defect state;
the detection sub-module is used for acquiring second real-time battery surface image data of the target battery after being overturned by using the chip image module, and inputting the second real-time battery surface image data into a target YOLOv5 battery surface detection model for detection to obtain a second target battery surface defect state;
and the removing sub-module is used for removing the target battery through the chip control module if the surface defect state of the second target battery is a general defect state.
8. The chip production control system for detecting a battery based on visual analysis according to claim 1, further comprising the steps of:
acquiring historical battery surface image data corresponding to a battery in a system, and performing data preprocessing on the historical battery surface image data to obtain battery surface image data to be trained;
a YOLOv5 battery surface detection model is established through a YOLO target detection algorithm, a Ghost Block light module is utilized to replace a C3 structure of the YOLOv5 battery surface detection model, a CBAM attention mechanism is introduced, and an initial YOLOv5 battery surface detection model is obtained;
inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training to obtain a target Yolov5 battery surface detection model, wherein the target Yolov5 battery surface detection model is arranged in a chip image module;
acquiring real-time battery surface image data of a target battery in a production line, and inputting the real-time battery surface image data into a target YOLOv5 battery surface detection model for identification to obtain a target battery surface defect state;
if the surface defect state of the target battery is a serious defect state, acquiring the position information of the target battery through a chip positioning module, and controlling the target battery through a chip control module;
if the surface defect state of the target battery is a general defect state, the target battery is subjected to overturn control by a chip control module, and the chip image module is used for detecting the surface defect of the target battery.
9. The chip production control system for detecting a battery based on visual analysis according to claim 1, further comprising the steps of:
building a YOLOv5 battery surface detection model through a YOLO target detection algorithm;
replacing a C3 structure of the YOLOv5 battery surface detection model by using a Ghost Block light module;
reducing parameters of the YOLOv5 battery surface detection model by decomposing standard convolution to generate an inexpensive feature map;
introducing a group of detection anchor frames on a low-level feature map of the Yolov5 battery surface detection model;
the decoupling detection head is utilized to independently process the regression and classification of the anchor frame to the Yolov5 battery surface detection model;
introducing a CBAM attention mechanism to the YOLOv5 battery surface detection model to obtain an initial YOLOv5 battery surface detection model.
10. The chip production control system for detecting a battery based on visual analysis according to claim 1, further comprising the steps of:
inputting the image data of the battery surface to be trained into the initial Yolov5 battery surface detection model for training;
adjusting the learning rate of the initial YOLOv5 battery surface detection model by using a cosine annealing algorithm, wherein the initial learning rate=0.01 and the final learning rate=0.1;
setting the threshold confidence coefficient=0.25 and NMS threshold value=0.45 of the initial Yolov5 battery surface detection model to obtain a target Yolov5 battery surface detection model;
and setting the target YOLOv5 battery surface detection model in a chip image module for detecting battery surface image data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311376720.3A CN117115162A (en) | 2023-10-24 | 2023-10-24 | Chip production control system for detecting battery based on visual analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311376720.3A CN117115162A (en) | 2023-10-24 | 2023-10-24 | Chip production control system for detecting battery based on visual analysis |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117115162A true CN117115162A (en) | 2023-11-24 |
Family
ID=88811393
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311376720.3A Pending CN117115162A (en) | 2023-10-24 | 2023-10-24 | Chip production control system for detecting battery based on visual analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117115162A (en) |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020042741A1 (en) * | 2018-08-27 | 2020-03-05 | 北京百度网讯科技有限公司 | Battery detection method and device |
| WO2022053001A1 (en) * | 2020-09-10 | 2022-03-17 | 上海航天精密机械研究所 | Weld seam internal defect intelligent detection device and method, and medium |
| CN114972316A (en) * | 2022-06-22 | 2022-08-30 | 合肥工业大学 | Battery case end surface defect real-time detection method based on improved YOLOv5 |
| CN115049611A (en) * | 2022-06-14 | 2022-09-13 | 东北大学 | Continuous casting billet crack defect identification method based on improved yolov5 |
| CN115170550A (en) * | 2022-07-29 | 2022-10-11 | 江苏大学 | Deep learning-based battery defect detection method and system |
| CN115187568A (en) * | 2022-07-26 | 2022-10-14 | 上海电力大学 | Power switch cabinet state detection method and system |
| CN115360798A (en) * | 2022-10-19 | 2022-11-18 | 中安芯界控股集团有限公司 | Online balancing method for battery clusters in battery energy storage system |
| CN115389965A (en) * | 2022-10-27 | 2022-11-25 | 中安芯界控股集团有限公司 | Big data based battery safety performance testing system and method |
| CN116228740A (en) * | 2023-04-07 | 2023-06-06 | 河海大学 | Small sample chip appearance defect detection method and detection system based on improved YOLOv5 |
| US20230184927A1 (en) * | 2021-12-15 | 2023-06-15 | Anhui University | Contextual visual-based sar target detection method and apparatus, and storage medium |
| CN116342563A (en) * | 2023-03-31 | 2023-06-27 | 晶科能源(上饶)有限公司 | Battery string detection method and system |
| CN116452505A (en) * | 2023-03-15 | 2023-07-18 | 北京科技大学 | Continuous casting billet internal defect detection and rating method based on improved YOLOv5 |
| CN116912218A (en) * | 2023-07-21 | 2023-10-20 | 四川九洲电器集团有限责任公司 | Deep learning-based battery surface defect detection method and device |
-
2023
- 2023-10-24 CN CN202311376720.3A patent/CN117115162A/en active Pending
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020042741A1 (en) * | 2018-08-27 | 2020-03-05 | 北京百度网讯科技有限公司 | Battery detection method and device |
| WO2022053001A1 (en) * | 2020-09-10 | 2022-03-17 | 上海航天精密机械研究所 | Weld seam internal defect intelligent detection device and method, and medium |
| US20230184927A1 (en) * | 2021-12-15 | 2023-06-15 | Anhui University | Contextual visual-based sar target detection method and apparatus, and storage medium |
| CN115049611A (en) * | 2022-06-14 | 2022-09-13 | 东北大学 | Continuous casting billet crack defect identification method based on improved yolov5 |
| CN114972316A (en) * | 2022-06-22 | 2022-08-30 | 合肥工业大学 | Battery case end surface defect real-time detection method based on improved YOLOv5 |
| CN115187568A (en) * | 2022-07-26 | 2022-10-14 | 上海电力大学 | Power switch cabinet state detection method and system |
| CN115170550A (en) * | 2022-07-29 | 2022-10-11 | 江苏大学 | Deep learning-based battery defect detection method and system |
| CN115360798A (en) * | 2022-10-19 | 2022-11-18 | 中安芯界控股集团有限公司 | Online balancing method for battery clusters in battery energy storage system |
| CN115389965A (en) * | 2022-10-27 | 2022-11-25 | 中安芯界控股集团有限公司 | Big data based battery safety performance testing system and method |
| CN116452505A (en) * | 2023-03-15 | 2023-07-18 | 北京科技大学 | Continuous casting billet internal defect detection and rating method based on improved YOLOv5 |
| CN116342563A (en) * | 2023-03-31 | 2023-06-27 | 晶科能源(上饶)有限公司 | Battery string detection method and system |
| CN116228740A (en) * | 2023-04-07 | 2023-06-06 | 河海大学 | Small sample chip appearance defect detection method and detection system based on improved YOLOv5 |
| CN116912218A (en) * | 2023-07-21 | 2023-10-20 | 四川九洲电器集团有限责任公司 | Deep learning-based battery surface defect detection method and device |
Non-Patent Citations (2)
| Title |
|---|
| 朱超平;杨永斌;: "基于改进的Faster-RCNN模型的汽车轮毂表面缺陷在线检测算法研究", 表面技术, no. 06 * |
| 李世杰: "《基于深度学习的无人机图像轻量化目标检测算法》", 《万方论文》, pages 24 - 37 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN113239930B (en) | Glass paper defect identification method, system, device and storage medium | |
| CN105388162B (en) | Raw material silicon chip surface scratch detection method based on machine vision | |
| CN108562589A (en) | A method of magnetic circuit material surface defect is detected | |
| CN113487570B (en) | High-temperature continuous casting billet surface defect detection method based on improved yolov5x network model | |
| CN113506286A (en) | Microwave chip defect detection method based on small sample data set of YOLOv5 algorithm | |
| CN113627435B (en) | Defect detection and identification method and system for ceramic tile | |
| CN110349125A (en) | A kind of LED chip open defect detection method and system based on machine vision | |
| CN113744244B (en) | Online visual detection system for measuring distance from edge of lithium battery pole piece to edge of pole lug | |
| CN111833313B (en) | Industrial product surface defect detection method and system based on deep active learning | |
| CN113177924A (en) | Industrial production line product flaw detection method | |
| CN115880298A (en) | Glass surface defect detection method and system based on unsupervised pre-training | |
| CN110020691B (en) | Liquid crystal screen defect detection method based on convolutional neural network impedance type training | |
| CN111275700A (en) | Terminal defect detection method and system based on deep learning | |
| CN113487563B (en) | EL image-based self-adaptive detection method for hidden cracks of photovoltaic module | |
| CN115866502A (en) | Microphone part surface defect online detection process | |
| CN114627122A (en) | Defect detection method and device | |
| CN116258682A (en) | PCB solder paste defect detection method based on PSPNet and improved YOLOv7 | |
| CN112560779B (en) | Method and equipment for identifying overflow of feeding port and feeding control system of stirring station | |
| CN113936291A (en) | Aluminum template quality inspection and recovery method based on machine vision | |
| CN117862068A (en) | Defective product sorting program verification method, device and storage medium | |
| CN117115162A (en) | Chip production control system for detecting battery based on visual analysis | |
| CN112967224A (en) | Electronic circuit board detection system, method and medium based on artificial intelligence | |
| CN117269168A (en) | New energy automobile precision part surface defect detection device and detection method | |
| CN114937015B (en) | Intelligent visual identification method and system in manufacturing of lithium battery pole piece | |
| CN116385353A (en) | Camera module abnormality detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20231124 |
|
| RJ01 | Rejection of invention patent application after publication |