CN110595401A - Detection method for detecting four corners of battery by using X-ray - Google Patents
Detection method for detecting four corners of battery by using X-ray Download PDFInfo
- Publication number
- CN110595401A CN110595401A CN201910809240.9A CN201910809240A CN110595401A CN 110595401 A CN110595401 A CN 110595401A CN 201910809240 A CN201910809240 A CN 201910809240A CN 110595401 A CN110595401 A CN 110595401A
- Authority
- CN
- China
- Prior art keywords
- products
- pictures
- corners
- battery
- namely
- 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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B15/00—Measuring arrangements characterised by the use of electromagnetic waves or particle radiation, e.g. by the use of microwaves, X-rays, gamma rays or electrons
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
- G01B15/00—Measuring arrangements characterised by the use of electromagnetic waves or particle radiation, e.g. by the use of microwaves, X-rays, gamma rays or electrons
- G01B15/02—Measuring arrangements characterised by the use of electromagnetic waves or particle radiation, e.g. by the use of microwaves, X-rays, gamma rays or electrons for measuring thickness
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- 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
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0004—Industrial image inspection
- G06T7/0008—Industrial image inspection checking presence/absence
-
- 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
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10116—X-ray image
-
- 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]
Abstract
The invention discloses a detection method for detecting four corners of a battery by using X-ray, which comprises the following steps: s1: product collection, S2: product classification, S3: classification algorithm, S4: segmentation algorithm, S5: identification, S6: and (6) outputting. The beneficial effects are that: the invention uses X-ray to obtain the winding picture in the battery cell, combines a classification module to classify the winding battery cell, and a segmentation measurement module segments and measures the anode and cathode end points of the battery cell to obtain the detailed data such as the deviation value of each layer of the battery cell, thereby automatically judging the winding quality of the battery cell.
Description
Technical Field
The invention relates to the field of X-ray detection, in particular to a detection method for detecting four corners of a battery by using X-ray.
Background
The cell winding process belongs to an extremely important link in the whole battery production process. How to guarantee the quality of the winding process also becomes an important factor.
Before the electric core is wound, the electric core is in a long strip shape, the width direction is taken as the axial direction, the electric core is wound in the length direction, after the winding is finished, the width is the height of the electric core, however, the height is larger than the width due to the fact that the electric core is easy to deviate in the winding process, the difference between the height and the width is deviation amount, and when the deviation amount is larger than a certain value, the electric core is a defective product. In order to detect the amount of deviation, an X-ray detection system is used.
The conventional X-ray detection system basically processes pictures only by using an image algorithm. For the cells with a large number of winding layers, the contrast is usually not clear enough, so that the information detected by the ordinary system is rough, and the reliability is low. The present design was originally applied to this.
In a common X-ray detection system, an X-ray is directly used for acquiring pictures, and a battery cell winding quality identification result is returned through an image algorithm.
Disadvantages of the above design
1. The traditional image algorithm is used for extremely inaccurate detection data of the battery cell with a large number of winding layers, and the reliability is poor;
2. for pictures with different winding layers, a long time cost optimization algorithm is needed to adapt to a new picture, and the result still cannot reach a high precision level;
3. the detection accuracy of the battery core with a plurality of winding layers is low, so that manual reinspection is still needed, all process scenes cannot be covered, and the labor cost cannot be well reduced.
Disclosure of Invention
The invention aims to provide a detection method for detecting four corners of a battery by using X-ray, which can realize the function of automatically detecting the over change (deviation) of the four corners of the lithium battery by using the X-ray and detect a poor winding battery cell so as to solve the problems in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme:
a detection method for detecting four corners of a battery by using X-ray rays comprises the following specific steps:
s1: collecting products, namely collecting a certain amount of battery cell products by a technician, wherein the battery cell products contain qualified products and unqualified products;
s2: and (4) product classification, wherein the technical personnel classify all the products collected in the S1, and the qualified products are classified into one class, and the unqualified products are classified into another class.
S3: the classification algorithm comprises the steps that four-corner information of the product of S2 is collected to a neural network for training, and then when the classification algorithm is used, obvious unqualified products are marked to form primary-selected qualified products and unqualified products, so that the obvious unqualified products are conveniently screened out;
s4: the segmentation algorithm is used for carrying out segmentation calculation on the qualified products primarily selected in the S3, the final output y ^ n is a probability matrix with the size equal to that of the input picture, and probability values are identified;
s5: marking, namely directly marking the products as qualified products and unqualified products according to the probability value of the S4 segmentation algorithm;
s6: and outputting, and displaying the identification result of the S5 so as to sort out qualified products and unqualified products.
The further technical scheme is as follows: the S3: the classification algorithm specifically comprises the following steps:
s31, acquiring pictures, namely shooting four corners of the battery by using X-ray to form the pictures, and acquiring information of the shot pictures;
s32, a priori type, whether the picture information of S31 is qualified or not is added;
s33, training the neural network, inputting the picture information with the mark into the neural network, and carrying out operations from S31 to S33 on a certain number of pictures;
s34, picture input, namely shooting four corners of the battery cell into pictures and inputting the pictures into the S35 part;
and S35, performing classification calculation, namely performing classification calculation on the picture information, wherein the calculation formula is as follows:
where n represents the number of layers in the network, y represents the final result after the activation function, σ represents the activation function, z is the convolution output of the current layer, ω is the weight matrix, and b is the bias matrix.
And S36, classifying the output, wherein the y ^ n of the final output is a probability value according to the calculation result of the S35.
The further technical scheme is as follows: the S4: the segmentation algorithm specifically comprises the following steps:
s41, acquiring pictures, namely shooting four corners of the battery by using X-ray to form the pictures, and acquiring information of the shot pictures;
s42, identifying the picture manually, and adding an identification whether the picture information of S41 is qualified;
s43, training the neural network, inputting the picture information with the mark into the neural network, and carrying out operations from S41 to S43 on a certain number of pictures;
s44, picture input, namely shooting four corners of the battery cell into pictures and inputting the pictures into the S45 part;
s45, segmentation calculation, namely, classification calculation is carried out on the picture information, and a calculation formula is as follows:
where n represents the number of layers in the network, y represents the final result after the activation function, σ represents the activation function, z is the convolution output of the current layer, ω is the weight matrix, and b is the bias matrix.
And S46, splitting the output, wherein the final output y ^ n is a probability matrix with the same size as the input picture according to the calculation result of S45.
The beneficial effects are that: the invention uses X-ray to obtain the winding picture in the battery cell, combines a classification module to classify the winding battery cell, and a segmentation measurement module segments and measures the anode and cathode end points of the battery cell to obtain the detailed data such as the deviation value of each layer of the battery cell, thereby automatically judging the winding quality of the battery cell.
The design is added with an image classification module on the basis of an X-ray detection system, can quickly and accurately classify the winding quality of the battery cell, and is also added with an image measurement module, so that the winding images of the battery cells of various levels are accurately measured, and accurate and detailed data are obtained. Therefore, the quality of the core film can be intelligently judged, the core film has extremely high accuracy, and meanwhile, the core films with different winding layers can be well compatible, and the core film does not need to be adapted to different conditions for a long time.
Drawings
FIG. 1: the overall flow chart of the invention.
FIG. 2: the invention relates to a classification algorithm flow chart.
FIG. 3: the invention relates to a segmentation algorithm flow chart.
Detailed Description
The technical solution in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.
Referring to fig. 1, a method for detecting four corners of a battery by using X-ray includes the following steps:
s1: collecting products, namely collecting a certain amount of battery cell products by a technician, wherein the battery cell products contain qualified products and unqualified products;
s2: and (4) product classification, wherein the technical personnel classify all the products collected in the S1, and the qualified products are classified into one class, and the unqualified products are classified into another class.
S3: the classification algorithm comprises the steps that four-corner information of the product of S2 is collected to a neural network for training, and then when the classification algorithm is used, obvious unqualified products are marked to form primary-selected qualified products and unqualified products, so that the obvious unqualified products are conveniently screened out;
s4: the segmentation algorithm is used for carrying out segmentation calculation on the qualified products primarily selected in the S3, the final output y ^ n is a probability matrix with the size equal to that of the input picture, and probability values are identified;
s5: marking, namely directly marking the products as qualified products and unqualified products according to the probability value of the S4 segmentation algorithm;
s6: and outputting, and displaying the identification result of the S5 so as to sort out qualified products and unqualified products.
As shown in fig. 2, a further technical solution: the S3: the classification algorithm specifically comprises the following steps:
s31, acquiring pictures, namely shooting four corners of the battery (qualified products or unqualified products) to form the pictures by using an X-ray, and acquiring information of the shot pictures;
s32, a priori classification, namely whether the picture information of the S31 is qualified or not is marked (for example, the picture information is marked as a qualified product or an unqualified product);
s33, training the neural network, inputting the picture information with the mark into the neural network, and carrying out operations from S31 to S33 on a certain number of pictures;
s34, picture input, (S31-S33 are learning stages, and S34-S36 are using stages, namely, whether a battery cell is a qualified product or an unqualified product can be identified, and certain accuracy can be achieved) four corners of the battery cell are shot into pictures, and the pictures are input to an S35 part;
and S35, performing classification calculation, namely performing classification calculation on the picture information, wherein the calculation formula is as follows:
where n represents the number of layers in the network, y represents the final result after the activation function, σ represents the activation function, z is the convolution output of the current layer, ω is the weight matrix, and b is the bias matrix.
And S36, classifying and outputting, wherein the y ^ n of the final output is a probability value (such as 98.5% of similarity with qualified products, 98.5% of similarity with unqualified products, and the like) according to the calculation result of S35.
As shown in fig. 3, a further technical solution: the S4: the segmentation algorithm specifically comprises the following steps:
s41, acquiring pictures, namely shooting four corners of the battery (the qualified products after primary selection) to form the pictures by using an X-ray, and acquiring information of the shot pictures (the qualified products after primary selection contain substantial unqualified products which can be identified by technicians of the unqualified products);
s42, identifying the picture manually, and adding an identification whether the picture is qualified or not (namely, the identification is qualified or unqualified in nature) to the picture information of S41;
s43, training the neural network, inputting the picture information with the mark into the neural network, and carrying out operations from S41 to S43 on a certain number of pictures;
s44, picture input, (S41-S43 are learning stages, and S44-S46 are using stages, namely, whether a battery cell is a qualified product or an unqualified product can be identified, and certain accuracy can be achieved) four corners of the battery cell are shot into pictures, and the pictures are input to an S45 part;
s45, segmentation calculation, namely, classification calculation is carried out on the picture information, and a calculation formula is as follows:
where n represents the number of layers in the network, y represents the final result after the activation function, σ represents the activation function, z is the convolution output of the current layer, ω is the weight matrix, and b is the bias matrix.
And S46, splitting the output, wherein the final output y ^ n is a probability matrix with the same size as the input picture according to the calculation result of S45.
It will be evident to those skilled in the art that the invention is not limited to the details of the foregoing illustrative embodiments, and that the present invention may be embodied in other specific forms without departing from the spirit or essential attributes thereof. The present embodiments are therefore to be considered in all respects as illustrative and not restrictive, the scope of the invention being indicated by the appended claims rather than by the foregoing description, and all changes which come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein. Any reference sign in a claim should not be construed as limiting the claim concerned.
Furthermore, it should be understood that although the present description refers to embodiments, not every embodiment may include only a single embodiment, and such description is for clarity only, and those skilled in the art will be able to make the description as a whole, and the embodiments may be appropriately combined to form other embodiments as will be apparent to those skilled in the art.
The invention discloses a detection method for detecting four corners of a battery by using X-ray, which comprises the following steps: s1: product collection, S2: product classification, S3: classification algorithm, S4: segmentation algorithm, S5: identification, S6: and (6) outputting. The beneficial effects are that: the invention uses X-ray to obtain the winding picture in the battery cell, combines a classification module to classify the winding battery cell, and a segmentation measurement module segments and measures the anode and cathode end points of the battery cell to obtain the detailed data such as the deviation value of each layer of the battery cell, thereby automatically judging the winding quality of the battery cell.
Claims (3)
1. A detection method for detecting four corners of a battery by using X-ray rays is characterized by comprising the following steps: the method comprises the following specific steps:
s1: collecting products, namely collecting a certain amount of battery cell products by a technician, wherein the battery cell products contain qualified products and unqualified products;
s2: classifying products, namely classifying all the products collected in the step S1 by a technician, wherein qualified products are classified into one class, and unqualified products are classified into another class;
s3: the classification algorithm comprises the steps that four-corner information of the product of S2 is collected to a neural network for training, and then when the classification algorithm is used, obvious unqualified products are marked to form primary-selected qualified products and unqualified products, so that the obvious unqualified products are conveniently screened out;
s4: the segmentation algorithm is used for carrying out segmentation calculation on the qualified products primarily selected in the S3, the final output y ^ n is a probability matrix with the size equal to that of the input picture, and probability values are identified;
s5: marking, namely directly marking the products as qualified products and unqualified products according to the probability value of the S4 segmentation algorithm;
s6: and outputting, and displaying the identification result of the S5 so as to sort out qualified products and unqualified products.
2. The method for detecting the four corners of the battery by using the X-ray as claimed in claim 1, wherein: the S3: the classification algorithm specifically comprises the following steps:
s31, acquiring pictures, namely shooting four corners of the battery by using X-ray to form the pictures, and acquiring information of the shot pictures;
s32, a priori type, whether the picture information of S31 is qualified or not is added;
s33, training the neural network, inputting the picture information with the mark into the neural network, and carrying out operations from S31 to S33 on a certain number of pictures;
s34, picture input, namely shooting four corners of the battery cell into pictures and inputting the pictures into the S35 part;
and S35, performing classification calculation, namely performing classification calculation on the picture information, wherein the calculation formula is as follows:
；
and S36, classifying the output, wherein the y ^ n of the final output is a probability value according to the calculation result of the S35.
3. The method for detecting the four corners of the battery by using the X-ray as claimed in claim 1, wherein: the S4: the segmentation algorithm specifically comprises the following steps:
s41, acquiring pictures, namely shooting four corners of the battery by using X-ray to form the pictures, and acquiring information of the shot pictures;
s42, identifying the picture manually, and adding an identification whether the picture information of S41 is qualified;
s43, training the neural network, inputting the picture information with the mark into the neural network, and carrying out operations from S41 to S43 on a certain number of pictures;
s44, picture input, namely shooting four corners of the battery cell into pictures and inputting the pictures into the S45 part;
s45, segmentation calculation, namely, classification calculation is carried out on the picture information, and a calculation formula is as follows:
；
and S46, splitting the output, wherein the final output y ^ n is a probability matrix with the same size as the input picture according to the calculation result of S45.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910809240.9A CN110595401A (en) | 2019-08-29 | 2019-08-29 | Detection method for detecting four corners of battery by using X-ray |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910809240.9A CN110595401A (en) | 2019-08-29 | 2019-08-29 | Detection method for detecting four corners of battery by using X-ray |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110595401A true CN110595401A (en) | 2019-12-20 |
Family
ID=68856302
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910809240.9A Pending CN110595401A (en) | 2019-08-29 | 2019-08-29 | Detection method for detecting four corners of battery by using X-ray |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110595401A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112465814A (en) * | 2020-12-17 | 2021-03-09 | 无锡日联科技股份有限公司 | Battery overlap calculation method and device based on deep learning |
| CN113552148A (en) * | 2020-04-13 | 2021-10-26 | 东芝It·控制系统株式会社 | Nondestructive inspection device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN207636021U (en) * | 2017-11-15 | 2018-07-20 | 宁德新能源科技有限公司 | Battery core detection device |
| CN108346137A (en) * | 2017-01-22 | 2018-07-31 | 上海金艺检测技术有限公司 | Defect inspection method for industrial x-ray weld image |
| CN108398084A (en) * | 2018-02-01 | 2018-08-14 | 深圳前海优容科技有限公司 | A kind of battery pole piece detection device, system, laminating machine and method |
| US10304208B1 (en) * | 2018-02-12 | 2019-05-28 | Avodah Labs, Inc. | Automated gesture identification using neural networks |
| CN109827973A (en) * | 2019-03-12 | 2019-05-31 | 东莞市骏智机电科技有限公司 | A kind of detection battery core protective film visible detection method |
| KR20190086338A (en) * | 2018-01-12 | 2019-07-22 | 한국과학기술원 | Method for processing x-ray computed tomography image using artificial neural network and apparatus therefor |
-
2019
- 2019-08-29 CN CN201910809240.9A patent/CN110595401A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108346137A (en) * | 2017-01-22 | 2018-07-31 | 上海金艺检测技术有限公司 | Defect inspection method for industrial x-ray weld image |
| CN207636021U (en) * | 2017-11-15 | 2018-07-20 | 宁德新能源科技有限公司 | Battery core detection device |
| KR20190086338A (en) * | 2018-01-12 | 2019-07-22 | 한국과학기술원 | Method for processing x-ray computed tomography image using artificial neural network and apparatus therefor |
| CN108398084A (en) * | 2018-02-01 | 2018-08-14 | 深圳前海优容科技有限公司 | A kind of battery pole piece detection device, system, laminating machine and method |
| US10304208B1 (en) * | 2018-02-12 | 2019-05-28 | Avodah Labs, Inc. | Automated gesture identification using neural networks |
| CN109827973A (en) * | 2019-03-12 | 2019-05-31 | 东莞市骏智机电科技有限公司 | A kind of detection battery core protective film visible detection method |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113552148A (en) * | 2020-04-13 | 2021-10-26 | 东芝It·控制系统株式会社 | Nondestructive inspection device |
| CN113552148B (en) * | 2020-04-13 | 2023-11-17 | 东芝It·控制系统株式会社 | Nondestructive inspection device |
| CN112465814A (en) * | 2020-12-17 | 2021-03-09 | 无锡日联科技股份有限公司 | Battery overlap calculation method and device based on deep learning |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115351598B (en) | Method for detecting bearing of numerical control machine tool | |
| US10853695B2 (en) | Method and system for cell annotation with adaptive incremental learning | |
| CN108520114B (en) | Textile fabric defect detection model and training method and application thereof | |
| CN113077453B (en) | Circuit board component defect detection method based on deep learning | |
| CN113674273B (en) | Optical detection method and system based on product defects and readable storage medium | |
| CN108090472B (en) | Pedestrian re-identification method and system based on multi-channel consistency characteristics | |
| US11790515B2 (en) | Detecting defects in semiconductor specimens using weak labeling | |
| WO2017092427A1 (en) | Electronic element positioning method and apparatus | |
| CN104952073B (en) | Scene Incision method based on deep learning | |
| CN111310756B (en) | Damaged corn particle detection and classification method based on deep learning | |
| CN113313712B (en) | Battery gluing defect detection method and device, electronic equipment and storage medium | |
| CN110595401A (en) | Detection method for detecting four corners of battery by using X-ray | |
| CN109800712A (en) | A kind of vehicle detection method of counting and equipment based on depth convolutional neural networks | |
| CN115170550A (en) | Deep learning-based battery defect detection method and system | |
| CN111652846B (en) | Semiconductor defect identification method based on characteristic pyramid convolution neural network | |
| CN111626989B (en) | High-precision detection network training method for lack-of-label pathological image | |
| CN115439411A (en) | Method and device for detecting polarity of circuit board component, medium and electronic equipment | |
| CN111832616A (en) | Method and system for identifying airplane model by using remote sensing image of multiple types of depth maps | |
| CN114359619A (en) | Incremental learning-based power grid defect detection method, device, equipment and medium | |
| CN112419243B (en) | Power distribution room equipment fault identification method based on infrared image analysis | |
| CN115953371A (en) | Insulator defect detection method, device, equipment and storage medium | |
| CN116363573A (en) | Transformer substation equipment state anomaly identification method and system | |
| CN114418011B (en) | Method, equipment and system for analyzing bad causes of products and storage medium | |
| CN113837178A (en) | Deep learning-based automatic positioning and unified segmentation method for meter of transformer substation | |
| CN111369508A (en) | Defect detection method and system for metal three-dimensional lattice structure |
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: 20191220 |
|
| RJ01 | Rejection of invention patent application after publication |