Battery defect detection method, device and electronic device

Abstract

The application discloses a battery defect detection method and device and electronic equipment, and relates to the technical field of batteries. The method comprises the following steps: acquiring a first image and a second image obtained by shooting a battery, wherein the first image and the second image comprise images of different dimensions; performing registration processing on the second image according to the first image; and determining a defect detection result of the battery according to the first image and the second image after registration.

Description

Technical Field

[0001] This application belongs to the field of battery technology, and in particular relates to a battery defect detection method, apparatus and electronic device. Background Technology

[0002] With the rapid development of battery technology, batteries are being used in various industries, especially as a core component of new energy vehicles. Currently, people not only demand large capacity batteries, but also pay close attention to their quality. However, due to the complexity of battery manufacturing processes, errors can occur during production, leading to defects such as misalignment of components, external bulging, or cracking. While defect detection is performed during battery production, it often suffers from low accuracy. Summary of the Invention

[0003] The purpose of this application is to provide a battery defect detection method, apparatus, and electronic device that can solve the problem of low detection accuracy that is commonly found in current battery defect detection.

[0004] In a first aspect, embodiments of this application provide a battery defect detection method, including:

[0005] Acquire a first image and a second image obtained from shooting the battery, wherein the first image and the second image include images of different dimensions;

[0006] Based on the first image, the second image is registered.

[0007] Based on the first image and the registered second image, the defect detection result of the battery is determined.

[0008] In this embodiment, a first image and a second image of the battery from different dimensions are acquired. Then, the second image is registered based on the first image. Finally, the battery defect detection result is determined based on the first image and the registered second image. This allows for the combined use of registered image information from different dimensions of the battery, enabling defect detection and improving the accuracy of the determined defect detection result, thus enhancing the overall accuracy of battery defect detection.

[0009] In some implementations, both the first image and the second image include N preset calibration objects, where N is a positive integer.

[0010] The registration process of the second image based on the first image includes:

[0011] Based on the N calibration objects, N pairs of feature points are determined between the first image and the second image. The N pairs of feature points correspond one-to-one with the N preset calibration objects, and each pair of feature points includes the feature points of its corresponding preset calibration object in the first image and the second image.

[0012] The second image is registered based on the N pairs of feature points.

[0013] In this embodiment, N pairs of feature points are determined between the first image and the second image by using N preset calibration objects in the first image and the second image. This improves the efficiency of determining matching feature point pairs in the first image and the second image during the registration process, thereby improving the efficiency of battery defect detection.

[0014] In some implementations, the first image includes a two-dimensional image, and the second image includes a three-dimensional image.

[0015] In this embodiment, battery defects can be detected using two-dimensional and three-dimensional images, which not only ensures the accuracy of defect detection but also reduces computational complexity and improves the efficiency of defect detection.

[0016] In some implementations, determining N pairs of feature points between the first image and the second image based on the N calibrated objects includes:

[0017] Obtain the grayscale image of the two-dimensional image;

[0018] Obtain the brightness image and depth image of the three-dimensional image;

[0019] Based on the N calibration objects, N pairs of feature points are determined in the grayscale image and the brightness image;

[0020] The registration process for the second image based on the N pairs of feature points includes:

[0021] The depth image is registered based on the N pairs of feature points.

[0022] In this embodiment, by acquiring the grayscale image of a two-dimensional image and the brightness and depth images of a three-dimensional image, and then determining N pairs of feature points in the grayscale and brightness images based on N calibration objects, and finally performing registration processing on the depth image based on the N pairs of feature points, the amount of computation in the registration process can be reduced and the registration efficiency can be improved.

[0023] In some implementations, determining the defect detection result of the battery based on the first image and the registered second image includes:

[0024] Based on the grayscale image, at least one suspected defect area of ​​the battery in the two-dimensional image is determined;

[0025] Based on the registered depth image, determine the image region corresponding to each of the suspected defect regions in the three-dimensional image;

[0026] Based on the depth information of the image region corresponding to each suspected defect region, the defect detection sub-result for each suspected defect region is determined.

[0027] The defect detection results include defect detection sub-results corresponding to the at least one suspected defect area.

[0028] In this embodiment, suspected defect areas are first identified in the grayscale image, and then, based on the depth information of each suspected defect area in the depth image, defect detection sub-results for each suspected defect area are determined, thereby achieving battery defect detection. This reduces the computational load for identifying defect areas in the depth image, thus improving the efficiency of battery defect detection.

[0029] In some implementations, determining N pairs of feature points between the first image and the second image based on the N calibrated objects includes:

[0030] The first image and the second image are input into a preset prediction model, which then outputs N pairs of feature points between the first image and the second image.

[0031] The preset prediction model is trained using at least one image sample. Each image sample includes a historical third image and a fourth image, and N pairs of historical feature points between the third image and the fourth image. The N pairs of historical feature points include feature points pre-labeled in the third image and the fourth image based on the N preset calibration objects.

[0032] In this embodiment, by inputting the first image and the second image into a preset prediction model, the preset prediction model outputs N pairs of feature points between the first image and the second image, thereby improving the efficiency of determining N pairs of feature points between the first image and the second image, and thus improving the efficiency of battery defect detection.

[0033] In some embodiments, before acquiring the first and second images obtained from the camera battery, the process further includes:

[0034] Calibrate the intrinsic parameter matrix of a 2D camera.

[0035] The first image is an image captured by the two-dimensional camera based on the calibrated intrinsic parameter matrix.

[0036] In this embodiment, by calibrating the intrinsic parameter matrix of the two-dimensional camera and capturing a first image based on the calibrated intrinsic parameter matrix, the registration accuracy can be further improved, thereby improving the accuracy of battery defect detection.

[0037] Secondly, embodiments of this application provide a battery defect detection device, comprising:

[0038] The image acquisition module is used to acquire a first image and a second image obtained by shooting the battery, wherein the first image and the second image include images of different dimensions;

[0039] The registration module is used to perform registration processing on the second image based on the first image;

[0040] The defect detection module is used to determine the defect detection result of the battery based on the first image and the registered second image.

[0041] In this embodiment, a first image and a second image of the battery from different dimensions are acquired. Then, the second image is registered based on the first image. Finally, the battery defect detection result is determined based on the first image and the registered second image. In this way, by combining the registered image information of the battery from different dimensions, battery defect detection can be achieved, thereby improving the accuracy of the determined defect detection result, i.e., improving the accuracy of battery defect detection.

[0042] In some implementations, both the first image and the second image include N preset calibration objects, where N is a positive integer.

[0043] The registration module includes:

[0044] The feature point pair determination unit is used to determine N pairs of feature points between the first image and the second image based on the N calibration objects. The N pairs of feature points correspond one-to-one with the N preset calibration objects, and each feature point pair includes the feature points of its corresponding preset calibration object in the first image and the second image.

[0045] The registration unit is used to perform registration processing on the second image based on the N pairs of feature points.

[0046] In this embodiment, N pairs of feature points are determined between the first image and the second image by using N preset calibration objects in the first image and the second image. This improves the efficiency of determining matching feature point pairs in the first image and the second image during the registration process, thereby improving the efficiency of battery defect detection.

[0047] In some implementations, the first image includes a two-dimensional image, and the second image includes a three-dimensional image.

[0048] In this embodiment, battery defects can be detected using two-dimensional and three-dimensional images, which not only ensures the accuracy of defect detection but also reduces computational complexity and improves the efficiency of defect detection.

[0049] In some embodiments, the feature point pair determination unit includes:

[0050] A two-dimensional image processing subunit is used to acquire a grayscale image of the two-dimensional image;

[0051] A three-dimensional image processing subunit is used to acquire the brightness image and depth image of the three-dimensional image;

[0052] The feature point pair determination subunit is used to determine N pairs of feature points in the grayscale image and the brightness image based on the N calibration objects.

[0053] The registration unit is specifically used for:

[0054] The depth image is registered based on the N pairs of feature points.

[0055] In this embodiment, by acquiring the grayscale image of a two-dimensional image and the brightness and depth images of a three-dimensional image, and then determining N pairs of feature points in the grayscale and brightness images based on N calibration objects, and finally performing registration processing on the depth image based on the N pairs of feature points, the amount of computation in the registration process can be reduced and the registration efficiency can be improved.

[0056] In some embodiments, the defect detection module includes:

[0057] A suspected defect area determination unit is used to determine at least one suspected defect area of ​​the battery in the two-dimensional image based on the grayscale image.

[0058] An image region determination unit is used to determine the image region corresponding to each of the suspected defect regions in the three-dimensional image based on the registered depth image.

[0059] The defect sub-result determination unit is used to determine the defect detection sub-result for each of the suspected defect regions based on the depth information of the image region corresponding to each of the suspected defect regions.

[0060] The defect detection results include defect detection sub-results corresponding to the at least one suspected defect area.

[0061] In this embodiment, suspected defect areas are first identified in the grayscale image, and then, based on the depth information of each suspected defect area in the depth image, defect detection sub-results for each suspected defect area are determined, thereby achieving battery defect detection. This reduces the computational load for identifying defect areas in the depth image, thus improving the efficiency of battery defect detection.

[0062] In some implementations, the feature point pair determination unit is specifically used for:

[0063] The image information from the first image and the second image is input into a preset prediction model, which then outputs N pairs of feature points between the first image and the second image.

[0064] The preset prediction model is trained using at least one image sample. Each image sample includes a historical third image and a fourth image, and N pairs of historical feature points between the third image and the fourth image. The N pairs of historical feature points include feature points pre-labeled in the third image and the fourth image based on the N preset calibration objects.

[0065] In this embodiment, by inputting the first image and the second image into a preset prediction model, the preset prediction model outputs N pairs of feature points between the first image and the second image, thereby improving the efficiency of determining N pairs of feature points between the first image and the second image, and thus improving the efficiency of battery defect detection.

[0066] In some implementations, it also includes:

[0067] The calibration module is used to calibrate the intrinsic parameter matrix of the 2D camera.

[0068] The first image is an image captured by the two-dimensional camera based on the calibrated intrinsic parameter matrix.

[0069] In this embodiment, by calibrating the intrinsic parameter matrix of the two-dimensional camera and capturing a first image based on the calibrated intrinsic parameter matrix, the registration accuracy can be further improved, thereby improving the accuracy of battery defect detection.

[0070] Thirdly, embodiments of this application provide an electronic device including a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method described in the first aspect.

[0071] Fourthly, embodiments of this application provide a readable storage medium on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first aspect.

[0072] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this application more obvious and understandable, the following are specific embodiments of this application. Attached Figure Description

[0073] Figure 1 This is a schematic diagram of the battery defect detection system provided in this application;

[0074] Figure 2 This is a schematic flowchart of the battery defect detection method provided in the embodiments of this application;

[0075] Figure 3 This is a schematic diagram of the battery defect detection device provided in the embodiments of this application.

[0076] Figure 4 This is a schematic diagram of the hardware structure of the electronic device provided in the embodiments of this application. Detailed Implementation

[0077] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0078] The terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such use of data can be interchanged where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first," "second," etc., are generally of the same class and the number of objects is not limited; for example, a first object can be one or more. Furthermore, in the specification and claims, "and / or" indicates at least one of the connected objects, and the character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0079] During battery production, due to the complexity of the manufacturing process, errors can occur at various stages, such as misalignment of components, bulging, or cracking of the battery, resulting in defects. Currently, to detect battery defects, a monocular camera is typically installed on the production line to capture two-dimensional images of the battery, and the defective areas are identified based on the image information. However, the limited information content in a two-dimensional image can lead to misjudgments, thus reducing the accuracy of battery defect detection.

[0080] To improve the accuracy of battery defect detection, this application proposes a new battery defect detection method, apparatus, and electronic device.

[0081] Please see Figure 1 This is an architecture diagram of the battery defect detection system provided in the embodiments of this application. As shown in Figure 1, the battery defect detection system includes a first camera 11, a second camera 122, and an electronic device 13. The first camera 11 and the second camera 12 are disposed on the production process line of the battery 20, and the first camera 11 and the second camera 12 respectively capture a first image and a second image of the battery 20 on the process line, and transmit the first image and the second image to the electronic device 13. The first camera 11 and the second camera 12 are cameras that capture images in different dimensions.

[0082] The first image and the second image can be images captured at the same time or images captured at different times, for example... Figure 1 The first camera 11 and the second camera 12 shown acquire images of the same battery 20 at different times.

[0083] Please see Figure 2 This is a flowchart illustrating the battery defect detection method provided in this application embodiment, applied to the aforementioned electronic device 13. For example... Figure 2 As shown, the method includes:

[0084] Step 201: Acquire the first image and the second image obtained by shooting the battery. The first image and the second image include images of different dimensions.

[0085] Step 202: Based on the first image, perform registration processing on the second image;

[0086] Step 203: Determine the defect detection results of the battery based on the first image and the registered second image.

[0087] In this embodiment, a first image and a second image of the battery from different dimensions are acquired. Then, the second image is registered based on the first image. Finally, the battery defect detection result is determined based on the first image and the registered second image. This allows for the combined use of registered image information from different dimensions of the battery, enabling defect detection and improving the accuracy of the determined defect detection result, thus enhancing the overall accuracy of battery defect detection.

[0088] In step 201 above, during the battery production process, the electronic device can acquire a first image and a second image obtained by photographing the battery.

[0089] The aforementioned battery may refer to a single battery cell, or a battery module composed of multiple battery cells, etc.

[0090] For example, a two-dimensional camera (i.e., the first camera 11 mentioned above) and a four-dimensional camera (i.e., the second camera 12 mentioned above) can be set on the battery cell production line. The two-dimensional camera and the four-dimensional camera are used to capture the appearance of the battery cell to obtain a two-dimensional image (i.e., the first image) and a four-dimensional image (i.e., the second image).

[0091] The dimension of the first image can be greater than the dimension of the second image, or the dimension of the second image can be greater than the dimension of the first image; no limitation is imposed here.

[0092] In step 202 above, after the electronic device acquires the first image and the second image, the electronic device can register the second image based on the first image.

[0093] The above-mentioned registration of the second image based on the first image can be performed by first extracting features from the first and second images to obtain feature points in the first and second images respectively; then matching the feature points in the first and second images to obtain matching feature point pairs; then obtaining image space coordinate transformation parameters through the matching feature point pairs; and finally performing image registration using the coordinate transformation parameters.

[0094] The feature extraction of the first and second images described above can be performed by detecting feature points in the first and second images using a preset feature point detection algorithm. For example, feature points in the first and second images can be extracted using algorithms such as Scale Invariant Feature Transform (SIFT).

[0095] The above-mentioned matching of feature points in the first image and the second image to obtain matching feature point pairs can be achieved by using a preset feature point matching algorithm. For example, it can be achieved by using at least one of optical flow methods and local descriptor-based methods to find feature points in the second image that match feature points in the first image.

[0096] The above-mentioned method of obtaining image spatial coordinate transformation parameters by matching feature point pairs can be achieved by performing spatial coordinate transformation processing on the feature points in the feature point pairs and calculating the image spatial coordinate transformation parameters. The aforementioned spatial coordinate transformation processing can include at least one of point transformation, affine transformation (including at least one of scaling, rotation, and translation), and projection transformation.

[0097] In step 203 above, after the electronic device performs registration processing on the second image, the electronic device can determine the defect detection result of the battery based on the first image and the registered second image.

[0098] The above-mentioned method of determining the battery defect detection result based on the first image and the registered second image can be achieved by the electronic device first performing image recognition on the first image to determine whether there is a suspected defective image region in the first image; if there is at least one suspected defective image region in the first image, the electronic device then determines the corresponding image region in the registered second image based on the coordinates of each suspected defective image region; finally, defect detection is performed on the image regions corresponding to each suspected defective image region to obtain the defect detection sub-results corresponding to each suspected defective image region, thereby determining whether the battery has a defect through the defect detection sub-results corresponding to each suspected defective image region.

[0099] The aforementioned determination of whether there is a suspected defective image region in the first image may be to determine whether there is an image region with at least one of the following: misalignment, bulge, and crack.

[0100] The above-mentioned determination of the battery defect detection result based on the first image and the registered second image may involve identifying a first defect region that meets the first preset defect condition in the first image and a second defect region that meets the second preset defect condition in the second image, and matching the identified first and second defect regions to obtain a matched image region; finally, the battery defect detection result is determined based on the matched defect region.

[0101] In some implementations, both the first image and the second image include N preset calibration objects, where N is a positive integer.

[0102] The above-mentioned registration process of the second image based on the first image may include:

[0103] Based on N calibration objects, N pairs of feature points are determined between the first image and the second image. The N pairs of feature points correspond one-to-one with the N preset calibration objects, and each feature point pair includes the feature points of its corresponding preset calibration object in the first image and the second image.

[0104] The second image is registered based on N pairs of feature points.

[0105] In this embodiment, N pairs of feature points are determined between the first image and the second image by using N preset calibration objects in the first image and the second image. This improves the efficiency of determining matching feature point pairs in the first image and the second image during the registration process, thereby improving the efficiency of battery defect detection.

[0106] The aforementioned N preset calibration objects can be pre-set objects used as reference calibration objects, and the shape and position of these N preset calibration objects remain fixed during the battery production process.

[0107] For example, when the battery is a battery cell, in order to ensure the transfer of the battery cell on the production line, the battery cell is usually clamped by at least one clamping member and driven to move. Therefore, a calibration mark can be set on the at least one clamped cell to form at least one calibration mark (i.e., the above N preset calibration objects). The calibration mark can be a dot or the like.

[0108] The aforementioned N preset calibration objects can be a single calibration object or multiple calibration objects. Specifically, the aforementioned N preset calibration objects can include 3 preset calibration objects, thereby ensuring both registration accuracy and registration efficiency.

[0109] The above method of determining N pairs of feature points between the first image and the second image based on N calibration objects can be achieved by identifying image regions that match each calibration object in the first image and the second image as feature points, and using the feature points identified in the first image and the second image as feature point pairs corresponding to the calibration object.

[0110] The above-mentioned matching process for the first image based on N pairs of feature points can be achieved by obtaining image space coordinate transformation parameters through these N pairs of feature points, and then performing image registration using the coordinate transformation parameters.

[0111] In some implementations, the first image includes a two-dimensional image, and the second image includes a three-dimensional image.

[0112] In this embodiment, battery defects can be detected using two-dimensional and three-dimensional images, which not only ensures the accuracy of defect detection but also reduces computational complexity and improves the efficiency of defect detection.

[0113] In the case where the first image includes a two-dimensional image and the second image includes a three-dimensional image, the above-mentioned registration process of the second image based on the first image can be performed by directly determining N pairs of feature points corresponding to the N calibration objects in the grayscale image of the two-dimensional image and the depth image of the three-dimensional image based on the N calibration objects.

[0114] In some implementations, determining N pairs of feature points between the first image and the second image based on N calibration objects includes:

[0115] Obtain the grayscale image of a two-dimensional image;

[0116] Obtain the brightness and depth images of the 3D image;

[0117] Based on N calibrated objects, determine N pairs of feature points in the grayscale image and the brightness image.

[0118] The above-mentioned registration process for the second image based on N pairs of feature points can include:

[0119] The depth image is registered based on N pairs of feature points.

[0120] In this embodiment, by acquiring the grayscale image of a two-dimensional image and the brightness and depth images of a three-dimensional image, and then determining N pairs of feature points in the grayscale and brightness images based on N calibration objects, and finally performing registration processing on the depth image based on the N pairs of feature points, the amount of computation in the registration process can be reduced and the registration efficiency can be improved.

[0121] Since the coordinates of each pixel in the brightness image and the depth image of the three-dimensional image are consistent, the above-mentioned registration process of the depth image based on N pairs of feature points can be performed by obtaining the image space coordinate transformation parameters between the grayscale image and the depth image through the above-mentioned N pairs of feature points, and then using the coordinate transformation parameters to perform image registration of the depth image.

[0122] In the case of registering the depth image based on N pairs of feature points, the determination of the battery defect detection result based on the first image and the registered second image can be achieved by identifying a first defect region that meets the first preset defect condition in the grayscale image and a second defect region that meets the second preset defect condition in the depth image, and matching the identified first and second defect regions to obtain a matched image region; finally, the battery defect detection result is determined based on the matched defect region.

[0123] For example, if at least one first defect region is identified in a grayscale image and at least one second defect region is identified in a depth-of-field image, and if the same defect region exists in at least one first defect region and at least one second defect region, then the battery is determined to be defective, and the same defect region is identified.

[0124] In some implementations, the defect detection result of the battery is determined based on the first image and the registered second image, including:

[0125] Based on the grayscale image, identify at least one suspected defect area of ​​the battery in the two-dimensional image;

[0126] Based on the registered depth image, determine the image region corresponding to each suspected defect region in the 3D image;

[0127] Based on the depth information of the image region corresponding to each suspected defect region, the defect detection sub-result for each suspected defect region is determined.

[0128] The defect detection results may include at least one defect detection sub-result corresponding to a suspected defect area.

[0129] In this embodiment, suspected defect areas are first identified in the grayscale image, and then, based on the depth information of each suspected defect area in the depth image, defect detection sub-results for each suspected defect area are determined, thereby achieving battery defect detection. This reduces the computational load for identifying defect areas in the depth image, thus improving the efficiency of battery defect detection.

[0130] Since the coordinates of each pixel in the registered depth image are mapped to the pixels in the grayscale image, after the electronic device determines the coordinates of each suspected defective image region, the electronic device can obtain the coordinates of the pixels in each suspected defective image region, and determine the pixels in the depth image whose coordinates are mapped to the pixels in each suspected defective image region based on the coordinates of the pixels in each suspected defective image region, and determine the image region including the determined pixels as the image region corresponding to the suspected defective image region.

[0131] The above-mentioned defect detection of the image regions corresponding to each suspected defective image region, and the resulting defect detection sub-results, can be obtained by determining whether the depth information of the image regions corresponding to each suspected defective image region meets a preset condition. If the preset condition is met, the resulting defect detection sub-result indicates that the suspected defective image region is a defective image region; if the preset condition is not met, the resulting defect detection sub-result indicates that the suspected defective image region is a defect-free image region.

[0132] For example, it can be determined whether the average depth value in the depth information of the image region corresponding to each suspected defective image region is greater than or equal to a preset depth value. If so, the suspected defective image region is determined to be an image region with defects, that is, the above defect detection sub-result indicates that the suspected defective image region is an image region with defects; otherwise, the suspected defective image region is determined to be an image region without defects, that is, the above defect detection sub-result indicates that the suspected defective image region is an image region without defects.

[0133] The above method of determining whether a battery is defective by identifying the defect detection sub-results corresponding to each suspected defective image region can be based on factors such as the area and number of image regions in the defect detection sub-results of at least one suspected defective image region.

[0134] For example, if the number of image regions indicating that the corresponding suspected defective image region is defective is greater than a preset number in the defect detection sub-result of at least one suspected defective image region, it can be determined whether the battery is defective.

[0135] In some implementations, determining N pairs of feature points between the first image and the second image based on N calibrated objects includes:

[0136] The first image and the second image are input into a preset prediction model, and the preset prediction model outputs N pairs of feature points between the first image and the second image.

[0137] The preset prediction model is trained using at least one image sample. Each image sample includes a historical third image and a fourth image, as well as N pairs of historical feature points between the third image and the fourth image. The N pairs of historical feature points include feature points pre-labeled in the third image and the fourth image based on N preset calibration objects.

[0138] In this embodiment, by inputting the first image and the second image into a preset prediction model, the preset prediction model outputs N pairs of feature points between the first image and the second image, thereby improving the efficiency of determining N pairs of feature points between the first image and the second image, and thus improving the efficiency of battery defect detection.

[0139] The aforementioned preset prediction model can be pre-trained using at least one of the aforementioned image training samples. That is, before inputting the first and second images into the preset prediction model, the process further includes: inputting at least one of the aforementioned image training samples into the deep neural model, and iteratively updating the network parameters of the deep neural model using the at least one image training sample to obtain the aforementioned preset prediction model. Since the process of training a deep neural model using training samples is known, it will not be elaborated upon here.

[0140] The third image mentioned above can be an image corresponding to the first image mentioned above, that is, the third image and the first image mentioned above are images with the same dimensional image data; similarly, the fourth image mentioned above is an image corresponding to the second image mentioned above.

[0141] The aforementioned N pairs of historical feature points can be pre-annotated manually in the aforementioned third and fourth images, corresponding to N preset calibration objects.

[0142] In some embodiments, before acquiring the first and second images obtained by the camera battery, the method further includes:

[0143] Calibrate the intrinsic parameter matrix of a 2D camera.

[0144] The first image is an image captured by a 2D camera based on the calibrated intrinsic parameter matrix.

[0145] In this embodiment, by calibrating the intrinsic parameter matrix of the two-dimensional camera and capturing a first image based on the calibrated intrinsic parameter matrix, the registration accuracy can be further improved, thereby improving the accuracy of battery defect detection.

[0146] The intrinsic parameter matrix of the aforementioned calibrated 2D camera can be calculated by acquiring calibration board images under different postures based on Zhang Zhengyou's calibration method.

[0147] The aforementioned calibration of the intrinsic parameter matrix of the 2D camera can be performed each time the battery production line is started; or it can be performed when the shooting scene of the 2D camera changes, for example, when the 2D camera moves.

[0148] Please see Figure 3 This is a schematic diagram of the battery defect detection device provided in an embodiment of this application. Figure 3 As shown, the device 300 includes:

[0149] Image acquisition module 301 is used to acquire a first image and a second image obtained by shooting the battery, the first image and the second image including images of different dimensions;

[0150] The registration module 302 is used to perform registration processing on the second image based on the first image;

[0151] The defect detection module 303 is used to determine the defect detection result of the battery based on the first image and the registered second image.

[0152] In some implementations, both the first image and the second image include N preset calibration objects, where N is a positive integer.

[0153] Registration module 302 may include:

[0154] The feature point pair determination unit is used to determine N pairs of feature points between the first image and the second image based on N calibration objects. The N pairs of feature points correspond one-to-one with the N preset calibration objects, and each feature point pair includes the feature points of its corresponding preset calibration object in the first image and the second image.

[0155] The registration unit is used to perform registration processing on the second image based on N pairs of feature points.

[0156] In some implementations, the first image includes a two-dimensional image, and the second image includes a three-dimensional image.

[0157] In some implementations, the feature point pair determination unit includes:

[0158] A two-dimensional image processing subunit is used to acquire grayscale images of two-dimensional images;

[0159] The 3D image processing subunit is used to acquire the brightness and depth images of the 3D image.

[0160] The feature point pair determination subunit is used to determine N pairs of feature points in a grayscale image and a brightness image based on N calibration objects.

[0161] The registration unit is specifically used for:

[0162] The depth image is registered based on N pairs of feature points.

[0163] In some implementations, the defect detection module 303 includes:

[0164] The suspected defect area determination unit is used to determine at least one suspected defect area of ​​the battery in a two-dimensional image based on the grayscale image.

[0165] The image region determination unit is used to determine the image region corresponding to each suspected defect region in the three-dimensional image based on the registered depth image;

[0166] The defect sub-result determination unit is used to determine the defect detection sub-result for each suspected defect region based on the depth information of the image region corresponding to each suspected defect region.

[0167] The defect detection results may include at least one defect detection sub-result corresponding to a suspected defect area.

[0168] In some implementations, the feature point pair determination unit is specifically used for:

[0169] The image information from the first image and the second image is input into a preset prediction model, and the preset prediction model outputs N pairs of feature points between the first image and the second image.

[0170] The preset prediction model can be trained by at least one image sample. Each image sample includes a historical third image and a fourth image, as well as N pairs of historical feature points between the third image and the fourth image. The N pairs of historical feature points include feature points pre-labeled in the third image and the fourth image based on N preset calibration objects.

[0171] In some implementations, it also includes:

[0172] The calibration module is used to calibrate the intrinsic parameter matrix of the 2D camera.

[0173] The first image is an image captured by a 2D camera based on the calibrated intrinsic parameter matrix.

[0174] Other details of the battery defect detection device according to embodiments of this application, in conjunction with the above. Figure 2 The battery defect detection method described in the example is similar and can achieve the same technical effect. For the sake of brevity, it will not be elaborated further here.

[0175] Figure 4 A schematic diagram of the hardware structure of the electronic device provided in an embodiment of this application is shown.

[0176] The electronic device may include a processor 401 and a memory 402 storing computer program instructions.

[0177] Specifically, the processor 401 may include a central processing unit (CPU), an application specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0178] Memory 402 may include mass storage for data or instructions. For example, and not limitingly, memory 402 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. In some instances, memory 402 may include removable or non-removable (or fixed) media, or memory 402 may be a non-volatile solid-state memory. In some embodiments, memory 402 may be internal or external to a battery device.

[0179] In some instances, memory 402 may be read-only memory (ROM). In one instance, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically rewritable ROM (EAROM), or flash memory, or a combination of two or more of these.

[0180] Memory 402 may include read-only memory (ROM), random access memory (RAM), disk storage media device, optical storage media device, flash memory device, electrical, optical, or other physical / tangible memory storage device. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to one aspect of this disclosure.

[0181] The processor 401 reads and executes computer program instructions stored in the memory 402 to achieve... Figure 2 The method in the illustrated embodiment achieves... Figure 2 The technical effects achieved by executing the methods / steps shown in the examples are not elaborated here for the sake of brevity.

[0182] In one example, the electronic device may also include a communication interface 403 and a bus 404. For example, Figure 4 As shown, the processor 401, memory 402, and communication interface 403 are connected through bus 404 and complete communication with each other.

[0183] The communication interface 403 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0184] Bus 404 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not as a limitation, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Extended Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), a Hyper Transport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 404 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, this application contemplates any suitable bus or interconnect.

[0185] The electronic device can perform the battery defect detection method in the embodiments of this application, thereby achieving a combination Figure 2 and Figure 3 A method and apparatus for detecting battery defects are described.

[0186] Furthermore, in conjunction with the battery defect detection method and apparatus in the above embodiments, this application embodiment can provide a computer storage medium for implementation. This computer storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the batteries and their control methods in the above embodiments.

[0187] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0188] The functional blocks shown in the above block diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0189] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0190] The aspects of this disclosure have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatuses, devices, and computer program products according to embodiments of this disclosure. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowchart illustrations, and combinations of blocks in the block diagrams and / or flowchart illustrations, can also be implemented by special-purpose hardware performing the specified functions or actions, or can be implemented by a combination of special-purpose hardware and computer instructions.

[0191] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and not to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. These modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application, and they should all be covered within the scope of the claims and specification of this application. In particular, as long as there is no structural conflict, the various technical features mentioned in the embodiments can be combined in any way. This application is not limited to the specific embodiments disclosed herein, but includes all technical solutions falling within the scope of the claims.

Claims

1. A method for detecting battery defects, characterized in that, include: Acquire a first image and a second image obtained by shooting the battery, the first image and the second image comprising images of different dimensions; the first image comprising a two-dimensional image and the second image comprising a three-dimensional image; Based on the first image, the second image is registered. Based on the first image and the registered second image, the defect detection result of the battery is determined; The step of determining the defect detection result of the battery based on the first image and the registered second image includes: Based on the first image, determine whether there is a suspected defective image region in the first image; If at least one suspected defective image region exists in the first image, the image region corresponding to each suspected defective image region is determined in the registered second image; Based on the image region corresponding to each suspected defective image region, a defect detection sub-result for each suspected defective image region is determined. The defect detection result includes the defect detection sub-result corresponding to the at least one suspected defective image region.

2. The method according to claim 1, characterized in that, Both the first image and the second image include N preset calibration objects, where N is a positive integer. The registration process of the second image based on the first image includes: Based on the N preset calibration objects, N pairs of feature points are determined between the first image and the second image. The N pairs of feature points correspond one-to-one with the N preset calibration objects, and each pair of feature points includes the feature points of its corresponding preset calibration object in the first image and the second image. The second image is registered based on the N pairs of feature points.

3. The method according to claim 2, characterized in that, The step of determining N pairs of feature points between the first image and the second image based on the N preset calibration objects includes: Obtain the grayscale image of the two-dimensional image; Obtain the brightness image and depth image of the three-dimensional image; Based on the N preset calibration objects, N pairs of feature points are determined in the grayscale image and the brightness image; The registration process for the second image based on the N pairs of feature points includes: The depth image is registered based on the N pairs of feature points.

4. The method according to claim 3, characterized in that, The step of determining the defect detection result of the battery based on the first image and the registered second image includes: Based on the grayscale image, at least one suspected defect area of ​​the battery in the two-dimensional image is determined; Based on the registered depth image, determine the image region corresponding to each of the suspected defect regions in the three-dimensional image; Based on the depth information of the image region corresponding to each of the suspected defect regions, the defect detection sub-results for each of the suspected defect regions are determined.

5. The method according to claim 2, characterized in that, The step of determining N pairs of feature points between the first image and the second image based on the N preset calibration objects includes: The first image and the second image are input into a preset prediction model, which then outputs N pairs of feature points between the first image and the second image. The preset prediction model is trained using at least one image sample. Each image sample includes a historical third image and a fourth image, and N pairs of historical feature points between the third image and the fourth image. The N pairs of historical feature points include feature points pre-labeled in the third image and the fourth image based on the N preset calibration objects.

6. The method according to claim 1, characterized in that, Before acquiring the first and second images obtained from the camera battery, the process also includes: Calibrate the intrinsic parameter matrix of a 2D camera. The first image is an image captured by the two-dimensional camera based on the calibrated intrinsic parameter matrix.

7. A battery defect detection device, characterized in that, include: The image acquisition module is used to acquire a first image and a second image obtained by shooting the battery, wherein the first image and the second image include images of different dimensions; the first image includes a two-dimensional image and the second image includes a three-dimensional image; The registration module is used to perform registration processing on the second image based on the first image; The defect detection module is used to determine the defect detection result of the battery based on the first image and the registered second image; The defect detection module is specifically used for: Based on the first image, determine whether there is a suspected defective image region in the first image; If at least one suspected defective image region exists in the first image, the image region corresponding to each suspected defective image region is determined in the registered second image; Based on the image region corresponding to each suspected defective image region, a defect detection sub-result for each suspected defective image region is determined. The defect detection result includes the defect detection sub-result corresponding to the at least one suspected defective image region.

8. The apparatus according to claim 7, characterized in that, Both the first image and the second image include N preset calibration objects, where N is a positive integer. The registration module includes: The feature point pair determination unit is used to determine N pairs of feature points between the first image and the second image based on the N preset calibration objects. The N pairs of feature points correspond one-to-one with the N preset calibration objects, and each feature point pair includes the feature points of its corresponding preset calibration object in the first image and the second image. The registration unit is used to perform registration processing on the second image based on the N pairs of feature points.

9. The apparatus according to claim 8, characterized in that, The feature point pair determination unit includes: A two-dimensional image processing subunit is used to acquire a grayscale image of the two-dimensional image; A three-dimensional image processing subunit is used to acquire the brightness image and depth image of the three-dimensional image; The feature point pair determination subunit is used to determine N pairs of feature points in the grayscale image and the brightness image based on the N preset calibration objects. The registration unit is specifically used for: The depth image is registered based on the N pairs of feature points.

10. The apparatus according to claim 9, characterized in that, The defect detection module includes: A suspected defect area determination unit is used to determine at least one suspected defect area of ​​the battery in the two-dimensional image based on the grayscale image. An image region determination unit is used to determine the image region corresponding to each of the suspected defect regions in the three-dimensional image based on the registered depth image. The defect sub-result determination unit is used to determine the defect detection sub-result of each suspected defect region based on the depth information of the image region corresponding to each suspected defect region.

11. The apparatus according to claim 8, characterized in that, The feature point pair determination unit is specifically used for: The image information from the first image and the second image is input into a preset prediction model, which then outputs N pairs of feature points between the first image and the second image. The preset prediction model is trained using at least one image sample. Each image sample includes a historical third image and a fourth image, and N pairs of historical feature points between the third image and the fourth image. The N pairs of historical feature points include feature points pre-labeled in the third image and the fourth image based on the N preset calibration objects.

12. The apparatus according to claim 7, characterized in that, Also includes: The calibration module is used to calibrate the intrinsic parameter matrix of the 2D camera. The first image is an image captured by the two-dimensional camera based on the calibrated intrinsic parameter matrix.

13. An electronic device, characterized in that, It includes a processor, a memory, and a program or instructions stored in the memory and executable on the processor, wherein the program or instructions, when executed by the processor, implement the steps of the method as described in any one of claims 1-6.

14. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method as described in any one of claims 1-6.

Images