A battery surface defect identification method, system, terminal and storage medium

Abstract

The present application discloses a battery surface defect recognition method, system, terminal and storage medium, the method comprising: obtaining an overall image of the battery surface, identifying defects in the overall image by image processing technology, and outputting a first defect recognition result; wherein the first defect recognition result includes the location information of the defect; based on the first defect recognition result, marking the area of ​​the battery surface to be recognized for a second time, and obtaining a local image of the battery surface corresponding to the area to be recognized for a second time, identifying defects in the local image by image processing technology, and outputting a second defect recognition result; the second defect recognition result includes the location, type, area and depth information of the defect; and using the second defect recognition result as the defect recognition result of the battery. The present application solves the problem that the accuracy and efficiency of battery surface defect recognition in the prior art cannot meet the requirements.

Description

Technical Field

[0001] The present application relates to the field of defect identification technology, and in particular to a battery surface defect identification method, system, terminal and storage medium. Background Art

[0002] With the continuous advancement and development of new energy technologies, batteries are increasingly used in new energy vehicles, energy storage power stations, aerospace and other fields. However, battery surface defects have an important impact on battery performance, reliability and safety. For example, when batteries are used in the field of new energy vehicles, due to the high energy density of new energy vehicle batteries, tiny cracks on the battery surface may also affect the vehicle's range and even cause safety hazards; therefore, battery surface defect identification has become a key link.

[0003] At present, although there are some battery surface defect recognition technologies, there are still deficiencies in recognition accuracy and efficiency. Traditional battery surface defect recognition methods mainly rely on manual visual inspection or simple automated equipment detection. These methods are limited in defect recognition speed and accuracy, and are not effective in identifying imperceptible defects such as tiny cracks. They cannot meet the high reliability requirements of batteries in new energy vehicles, energy storage power stations, aerospace and other fields. Summary of the invention

[0004] In order to solve the problem that the accuracy and efficiency of battery surface defect recognition in the prior art cannot meet the requirements, the present application provides a battery surface defect recognition method, system, terminal and storage medium.

[0005] In a first aspect, the present application provides a battery surface defect recognition method, which adopts the following technical solution: the method comprises: obtaining an overall picture of the battery surface, identifying defects in the overall picture by image processing technology, and outputting a first defect recognition result; wherein the first defect recognition result includes the location information of the defect;

[0006] Based on the first defect recognition result, mark the area to be recognized for secondary recognition on the battery surface, obtain a local image of the battery surface corresponding to the area to be recognized for secondary recognition, recognize defects in the local image by image processing technology, and output a second defect recognition result; the second defect recognition result includes the location, type, area and depth information of the defect;

[0007] The second defect recognition result is used as the defect recognition result of the battery.

[0008] By adopting the above-mentioned technical scheme, the present application adopts a phased detection method, combining the analysis of the overall and local images, first preliminarily identifying and quickly screening out most of the defects, and marking the defect locations, dividing the area for secondary identification according to the distribution of these defects, and further analyzing the local shots, the type, depth, size, location and other information of the defects can be accurately obtained, and the accuracy of defect identification is high, providing a reliable basis for the subsequent judgment of whether the product is qualified; and it avoids unnecessary invalid detection, saves time, and improves the efficiency of defect identification.

[0009] In a specific implementation scheme, based on the first defect recognition result, marking the area on the battery surface to be re-recognized specifically includes:

[0010] Based on the first defect recognition result, marking each defect point on the battery surface;

[0011] Based on a set defect point classification rule, traverse all defect points on the battery surface and classify each defect point as a first defect point or a second defect point;

[0012] Based on the distribution of the first defect points, a first identification area on the battery surface is marked; based on the distribution of the second defect points, a second identification area on the battery surface is marked; wherein one second defect point corresponds to one second identification area;

[0013] The area to be re-identified is constructed based on the first identification area and the second identification area.

[0014] By adopting the above technical solution, the defect points are first marked, and then each defect point is distinguished as a first defect point or a second defect point. After that, different methods are adopted for different types of defect points to obtain a first identification area and a second identification area, thereby forming an area on the battery surface to be identified for the second time.

[0015] In a specific implementation scheme, the defect point division rule includes:

[0016] For each defect point on the battery surface, count the number of defect points n on the battery surface whose distance from the defect point does not exceed the preset neighborhood radius; determine whether the number of defect points n exceeds the preset threshold value N, if so, mark the defect point as the first defect point, if not, mark the defect point as the second defect point.

[0017] By adopting the above technical solution, based on the defect point division rule, the defect points with relatively dense distribution are marked as first defect points, and the defect points with relatively sparse distribution are marked as second defect points.

[0018] In a specific possible implementation scheme, based on the distribution of the first defect points, marking a first identification area on the battery surface specifically includes:

[0019] For the first defect point, marking a first identification sub-region corresponding to the first defect point based on a preset neighborhood radius;

[0020] Determine whether there is overlap between the first identification sub-regions corresponding to the first defect points; if so, merge the overlapping first identification sub-regions into one first identification region; if not, mark the first identification sub-region separately as one first identification region.

[0021] By adopting the above technical solution, for each first defect point, its corresponding first identification sub-area is marked. If there is an overlap of the first identification sub-areas, it means that these defect points are distributed in a relatively close area. Therefore, the overlapping first identification sub-areas are merged. When performing secondary shooting and identification in the subsequent time, the number of secondary shooting and identification can be greatly reduced, avoiding unnecessary calculations.

[0022] In a specific possible implementation scheme, after merging the overlapping first identification sub-regions into one first identification region, the method further comprises:

[0023] Determine whether the area of ​​the first recognition area after merging exceeds a preset maximum area threshold;

[0024] If so, based on the maximum area threshold, the merged first recognition area is divided to obtain multiple first recognition areas.

[0025] By adopting the above technical solution, if the area is too large, it is not conducive to accurate identification of defects. Therefore, when the area of ​​the area is too large, secondary division is performed.

[0026] In a specific possible implementation scheme, before obtaining the overall image of the battery surface, the method further includes:

[0027] Acquiring surface feature information of the battery; wherein the surface feature information includes one or more of reflective features, texture features, color features, and shape features;

[0028] Based on the correspondence between the preset battery surface feature information and the shooting conditions, the battery is photographed using the shooting conditions corresponding to the acquired battery surface feature information to obtain an overall picture of the battery surface.

[0029] By adopting the above technical solution, according to the surface characteristic information of the battery, when photographing the battery, the most suitable shooting conditions, that is, the selection of the camera and the selection of the light source, are selected, so as to obtain an overall picture of the battery surface with high clarity which is conducive to subsequent defect identification, thereby further improving the accuracy of defect identification.

[0030] In a specific implementation scheme, after taking the second defect recognition result as the defect recognition result of the battery, the method further includes:

[0031] Based on the defect identification result of the battery, classify the battery having the first characteristic defect on the battery surface as an unqualified battery;

[0032] The first characteristic defect includes one or more of a depth of the defect exceeding a depth threshold, an area of ​​the defect exceeding an area threshold, and a position of the defect being at a preset first position.

[0033] By adopting the above technical solution, after obtaining the defect identification result of the battery, the battery is classified as qualified or not according to the result, which saves the subsequent manual screening process of the staff.

[0034] In a second aspect, the present application provides a battery surface defect recognition system, applying the battery surface defect recognition method in the first aspect or any one of the embodiments of the first aspect, the system comprising a preliminary recognition module and a secondary recognition module;

[0035] The preliminary identification module is used to obtain an overall picture of the battery surface, identify defects in the overall picture through image processing technology, and output a first defect identification result; wherein the first defect identification result includes location information of the defect;

[0036] The secondary identification module is used to mark the area to be secondary identified on the battery surface based on the first defect identification result, and obtain a local image of the battery surface corresponding to the area to be secondary identified, identify defects in the local image through image processing technology, and output a second defect identification result, and use the second defect identification result as the defect identification result of the battery; wherein the second defect identification result includes the location, type, area and depth information of the defect.

[0037] In a third aspect, the present application provides a terminal comprising: a processor, a memory and a communication bus; the communication bus is used to realize connection and communication between the processor and the memory, and the processor is used to execute one or more programs stored in the memory to realize the battery surface defect identification method in the above-mentioned first aspect or any possible implementation scheme of the first aspect.

[0038] In a fourth aspect, the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, and when the instructions are executed, the battery surface defect identification method in the above-mentioned first aspect or any possible implementation scheme of the first aspect is executed.

[0039] In summary, the technical solution of this application includes at least the following beneficial technical effects:

[0040] 1. This application adopts a phased detection method, combined with the analysis of overall and local images. First, a preliminary identification is performed to quickly screen out most of the defects, and the defect locations are marked. The area to be recognized for secondary identification is divided according to the distribution of these defects. Further analysis of local shots can accurately obtain information such as the type, depth, size, and location of the defects. The accuracy of defect recognition is high, providing a reliable basis for subsequent judgment on whether the product is qualified; it also avoids unnecessary invalid detection, saves time, and improves the efficiency of defect recognition; it is a good solution to the problem that the accuracy and efficiency of battery surface defect recognition in the prior art cannot meet the needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a schematic diagram of the main process of the battery surface defect identification method in an embodiment of the present application;

[0042] Figure 2 It is a schematic diagram of the process of marking the area to be re-identified in an embodiment of the present application. DETAILED DESCRIPTION

[0043] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below with reference to the accompanying drawings.

[0044] The present application embodiment provides a battery surface defect identification method, such as Figure 1 As shown, the method includes steps S1-S4.

[0045] S1, obtaining an overall picture of the battery surface, identifying defects in the overall picture through image processing technology, and outputting a first defect identification result; the first identification result includes location information of the defect.

[0046] In this step, a camera is first used to capture an overall picture of the battery surface. Through image processing algorithms, such as edge detection or deep learning models, potential defects such as scratches and dents on the battery surface can be automatically marked.

[0047] In a possible implementation, in this step, before identifying defects in the overall image by image processing technology, the overall image may be preprocessed; the preprocessing includes one or more of image filtering, grayscale conversion, histogram equalization, and binarization. Through such preprocessing, the image quality is significantly improved, and the key features in the image are enhanced, thereby laying the foundation for the accurate identification of battery surface defects, which not only improves the recognition accuracy of subsequent algorithms, but also reduces the impact of environmental factors on the recognition results.

[0048] S2, based on the first defect recognition result, mark the area on the battery surface to be recognized for a second time, obtain a local image of the battery surface corresponding to the area to be recognized for a second time, identify defects in the local image through image processing technology, and output a second defect recognition result; the second defect recognition result includes the location, type, area and depth information of the defect.

[0049] After the potential defects are initially identified, they need to be identified in more detail. In this step, a higher-resolution camera can be used to take a second shot of the area to be re-identified, obtain a local high-definition picture, and use more sophisticated image processing technology to specifically diagnose the type, size, depth and other information of the defects.

[0050] In one possible implementation, refer to Figure 2 In step S2, based on the first defect recognition result, marking the area on the battery surface to be recognized again specifically includes:

[0051] S21: Mark each defect point on the battery surface based on the first defect identification result.

[0052] S22, based on the set defect point classification rule, traverse all defect points on the battery surface and classify each defect point into a first defect point or a second defect point, wherein the first defect point represents a densely distributed defect point, and the second defect point represents a sparsely distributed defect point.

[0053] In a possible implementation manner, the defect point division rule includes:

[0054] For each defect point on the battery surface, count the number of defect points n on the battery surface whose distance from the defect point does not exceed the preset neighborhood radius; determine whether the number of defect points n exceeds the preset threshold value N, if so, mark the defect point as the first defect point, if not, mark the defect point as the second defect point.

[0055] According to the defect point division rule, the defect points with relatively dense distribution are marked as the first defect points, and the defect points with relatively sparse distribution are marked as the second defect points.

[0056] S23, marking a first identification area on the battery surface based on the distribution of the first defect points;

[0057] Based on the distribution of the second defect points, a second identification area on the battery surface is marked; wherein one second defect point corresponds to one second identification area.

[0058] In a possible implementation, step S23, based on the distribution of the first defect points, marking a first identification area on the battery surface, specifically includes steps A1-A2.

[0059] A1, for the first defect point, marking a first identification sub-region corresponding to the first defect point based on a preset neighborhood radius;

[0060] A2, determining whether there is overlap between the first identification sub-regions corresponding to the first defect points; if so, merging the overlapping first identification sub-regions into one first identification region; if not, marking the first identification sub-regions separately as one first identification region.

[0061] Through steps A1-A2, for each first defect point, its corresponding first identification sub-area is marked. If there is overlap of the first identification sub-areas, it means that these defect points are distributed in a relatively close area. Therefore, the overlapping first identification sub-areas are merged. When performing secondary shooting and identification in the subsequent time, the number of secondary shooting and identification can be greatly reduced, avoiding unnecessary calculations.

[0062] In a possible implementation manner, after step A2, merging the overlapping second identification sub-regions into one first identification region, further includes the following steps:

[0063] A3, determining whether the area of ​​the first recognition area after merging exceeds a preset maximum area threshold;

[0064] A4: If yes, then based on the maximum area threshold, the merged first recognition area is divided to obtain multiple first recognition areas.

[0065] Through the above steps A3-A4, if the area is too large, it is not conducive to the accurate identification of defects. Therefore, when the area of ​​the area is too large, a secondary division is performed. This step can be divided according to the set single area area limit value, that is, the area of ​​each of the first identification areas does not exceed the single area area limit value. Technical personnel in this field can also divide according to other rules, and this application does not make specific limitations on this.

[0066] S24: construct the area to be recognized again based on the first recognition area and the second recognition area.

[0067] Through the above steps S21-S24, the defect points are first marked, and then each defect point is distinguished as a first defect point or a second defect point. After that, different methods are used for different types of defect points to obtain the first identification area and the second identification area, thereby forming the area to be recognized for the second time on the battery surface.

[0068] S3: Use the second defect recognition result as the battery defect recognition result.

[0069] In the above steps S1-S3, the present application adopts a phased detection method, combined with the analysis of the overall and local images, first preliminarily identifies and quickly screens out most of the defects, and marks the defect locations, and divides the area for secondary identification according to the distribution of these defects. Further analysis of local shots can accurately obtain information such as the type, depth, size, and location of the defects. The accuracy of defect identification is high, which provides a reliable basis for subsequent judgment on whether the product is qualified; it also avoids unnecessary invalid detection, saves time, and improves the efficiency of defect identification; it is a good solution to the problem that the accuracy and efficiency of battery surface defect identification in the prior art cannot meet the needs.

[0070] For example, in the initial identification stage, potential defects may only be marked with an accuracy rate of 80%, and the specific information of the defects is not accurate enough at this time. However, after secondary shooting and local identification, this accuracy rate will be effectively improved.

[0071] In a possible implementation manner, after step S3, using the second defect recognition result as the battery defect recognition result, the method further includes step S4.

[0072] S4, based on the defect identification result of the battery, classifying the battery having the first characteristic defect on the battery surface as an unqualified battery;

[0073] The first characteristic defect includes one or more of a depth of the defect exceeding a depth threshold, an area of ​​the defect exceeding an area threshold, and a position of the defect being at a preset first position.

[0074] Specifically, technicians in this field can flexibly set the first position according to the shape of the battery and other timing conditions. When the position of the defect is the first position, it means that the defect here, regardless of the area, depth or type, will have a significant impact on the characteristics of the battery. Therefore, when this happens, the battery needs to be classified as an unqualified battery to avoid problems when the battery is used in practice.

[0075] Through step S4, after the defect recognition result of the battery is obtained, the battery is classified as qualified or not according to the result, which saves the subsequent manual screening process of the staff.

[0076] In a possible implementation, before step S1, obtaining the overall image of the battery surface, further includes steps B1-B2.

[0077] B1, obtaining surface feature information of the battery; wherein the surface feature information includes one or more of reflective features, texture features, color features and shape features;

[0078] B2, based on the preset correspondence between the battery surface feature information and the shooting conditions, the battery is photographed using the shooting conditions corresponding to the acquired battery surface feature information to obtain an overall picture of the battery surface.

[0079] Through steps B1-B2, according to the surface characteristic information of the battery, when photographing the battery, the most suitable shooting conditions, that is, the selection of the camera and the selection of the light source, are selected, so as to obtain an overall picture of the battery surface with high clarity that is conducive to subsequent defect identification, thereby further improving the accuracy of defect identification.

[0080] In a possible implementation, the method further includes:

[0081] The defect identification results of the same batch of batteries are counted and analyzed to provide a reference for the adjustment of battery processing technology and / or processing parameters.

[0082] For example, if the same batch of batteries all have defects in the same position, or the defects of the batteries in this batch are significantly higher than normal, the staff can adjust the processing technology and parameters in a targeted manner to reduce the occurrence of defects.

[0083] The embodiment of the present application provides a battery surface defect recognition system, which applies the battery surface defect recognition method described in the above embodiment, and the system includes a preliminary recognition module and a secondary recognition module;

[0084] The preliminary identification module is used to obtain an overall picture of the battery surface, identify defects in the overall picture through image processing technology, and output a first defect identification result; wherein the first defect identification result includes location information of the defect;

[0085] The secondary identification module is used to mark the area to be secondary identified on the battery surface based on the first defect identification result, and obtain a local image of the battery surface corresponding to the area to be secondary identified, identify defects in the local image through image processing technology, and output a second defect identification result, and use the second defect identification result as the defect identification result of the battery; wherein the second defect identification result includes the location, type, area and depth information of the defect.

[0086] An embodiment of the present application provides a terminal, including: a processor, a memory and a communication bus; the communication bus is used to realize connection and communication between the processor and the memory, and the processor is used to execute one or more programs stored in the memory to implement the battery surface defect identification method as described in the above embodiment.

[0087] An embodiment of the present application provides a computer-readable storage medium, which stores instructions. When the instructions are executed, the battery surface defect identification method described in the above embodiment is executed.

[0088] The above are all preferred embodiments of the present application, and the protection scope of the present application is not limited thereto. Therefore, any equivalent changes made according to the structure, shape, and principle of the present application should be included in the protection scope of the present application.

Claims

1. A battery surface defect identification method, characterized in that: include: Taking an overall picture of the battery surface by a camera, identifying defects in the overall picture by image processing technology, and outputting a first defect identification result; wherein the first defect identification result includes location information of the defect; Based on the first defect recognition result, a battery surface area to be recognized for a second time is marked, and a local image of the battery surface corresponding to the area to be recognized for a second time is obtained by taking a second shot with a camera, and defects in the local image are recognized by image processing technology, and a second defect recognition result is output; the second defect recognition result includes information on the location, type, area, and depth of the defect; wherein the resolution of the camera used to obtain the local image is higher than the resolution of the camera used to obtain the overall image; Using the second defect recognition result as a defect recognition result of the battery; The step of marking the area on the battery surface to be re-identified based on the first defect identification result specifically includes: Based on the first defect recognition result, marking each defect point on the battery surface; Based on a set defect point classification rule, traverse all defect points on the battery surface and classify each defect point as a first defect point or a second defect point; Based on the distribution of the first defect points, a first identification area on the battery surface is marked; based on the distribution of the second defect points, a second identification area on the battery surface is marked; wherein one second defect point corresponds to one second identification area; Based on the first recognition area and the second recognition area, construct the area to be recognized again; The defect point division rules include: For each defect point on the battery surface, count the number of defect points n on the battery surface whose distance from the defect point does not exceed a preset neighborhood radius; determine whether the number of defect points n exceeds a preset threshold value N; if so, mark the defect point as a first defect point, if not, mark the defect point as a second defect point.

2. The battery surface defect identification method according to claim 1, characterized in that: Based on the distribution of the first defect points, marking a first identification area on the battery surface specifically includes: For the first defect point, marking a first identification sub-region corresponding to the first defect point based on a preset neighborhood radius; Determine whether there is overlap between the first identification sub-regions corresponding to the first defect points; if so, merge the overlapping first identification sub-regions into one first identification region; if not, mark the first identification sub-region separately as one first identification region.

3. The battery surface defect identification method according to claim 2, characterized in that: After merging the overlapping first recognition sub-regions into one first recognition region, the method further includes: Determine whether the area of ​​the first recognition area after merging exceeds a preset maximum area threshold; If so, based on the maximum area threshold, the merged first recognition area is divided to obtain multiple first recognition areas.

4. The battery surface defect identification method according to claim 1, characterized in that: Before getting a holistic picture of the battery surface: Acquiring surface feature information of the battery; wherein the surface feature information includes one or more of reflective features, texture features, color features, and shape features; Based on the correspondence between the preset battery surface feature information and the shooting conditions, the battery is photographed using the shooting conditions corresponding to the acquired battery surface feature information to obtain an overall picture of the battery surface.

5. The battery surface defect identification method according to claim 1, characterized in that: After using the second defect recognition result as the defect recognition result of the battery, the method further includes: Based on the defect identification result of the battery, classify the battery having the first characteristic defect on the battery surface as an unqualified battery; The first characteristic defect includes one or more of a depth of the defect exceeding a depth threshold, an area of ​​the defect exceeding an area threshold, and a position of the defect being at a preset first position.

6. A battery surface defect recognition system, characterized in that: It includes a preliminary recognition module and a secondary recognition module; The preliminary identification module is used to obtain an overall picture of the battery surface by taking pictures with a camera, identify defects in the overall picture by image processing technology, and output a first defect identification result; wherein the first defect identification result includes location information of the defect; The secondary recognition module is used to mark the area to be secondary recognized on the battery surface based on the first defect recognition result, and obtain a local image of the battery surface corresponding to the area to be secondary recognized by a camera for secondary shooting, identify defects in the local image by image processing technology, and output a second defect recognition result, and use the second defect recognition result as the defect recognition result of the battery; wherein the second defect recognition result includes the location, type, area and depth information of the defect; the resolution of the camera for obtaining the local image is higher than the resolution of the camera for obtaining the overall image; The secondary identification module marks the area on the battery surface to be secondary identified based on the first defect identification result, specifically including: Based on the first defect recognition result, marking each defect point on the battery surface; Based on a set defect point classification rule, traverse all defect points on the battery surface and classify each defect point as a first defect point or a second defect point; Based on the distribution of the first defect points, a first identification area on the battery surface is marked; based on the distribution of the second defect points, a second identification area on the battery surface is marked; wherein one second defect point corresponds to one second identification area; Based on the first recognition area and the second recognition area, construct the area to be recognized again; The defect point division rules include: For each defect point on the battery surface, count the number of defect points n on the battery surface whose distance from the defect point does not exceed a preset neighborhood radius; determine whether the number of defect points n exceeds a preset threshold value N; if so, mark the defect point as a first defect point, if not, mark the defect point as a second defect point.

7. A terminal, characterized in that: include: A processor, a memory and a communication bus; the communication bus is used to realize the connection and communication between the processor and the memory, and the processor is used to execute one or more programs stored in the memory to implement the battery surface defect identification method as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores instructions, and when the instructions are executed, the battery surface defect identification method according to any one of claims 1 to 5 is executed.

Images