Pin product deformation detection method by expanding visual field

Abstract

The invention discloses a pin product deformation detection method by expanding a visual field, and relates to the technical field of image processing, and the method comprises the steps: collecting an initial product image in an initial visual field range; dividing the product image to obtain a plurality of grid initial images in the plurality of grids, and performing deformation identification to obtain a plurality of pieces of initial deformation information; when deformation exists in the multiple pieces of initial deformation information, deformation trend prediction is carried out on the multiple edge grid initial images, and multiple pieces of deformation trend information are obtained; calculating to obtain a visual field expansion coefficient according to the multiple pieces of edge initial deformation information and the multiple pieces of deformation trend information, adjusting an initial visual field range, obtaining an expanded visual field range, continuing deformation detection of the target pin product, and obtaining a final deformation detection result according to the multiple pieces of initial deformation information. The technical problems that in the prior art, the deformation detection range is limited, and the detection precision is insufficient are solved, and the technical effect of improving the deformation detection precision and comprehensiveness of the target pin product is achieved.

Description

Technical Field

[0001] The invention relates to the technical field of image processing, and in particular to a method for detecting deformation of a pin product by expanding a field of view. Background Art

[0002] With the continuous development of modern manufacturing, especially in the production of precision electronic components and micro-devices, the quality control of pin products has become increasingly important. Traditional pin product deformation detection methods are usually limited by the detection range and field of view, making it difficult to fully and accurately capture the deformation of the product. This detection method often only covers the central part of the product, and the edges and hard-to-reach areas are easily overlooked, resulting in incomplete deformation detection, which in turn affects product quality control. Summary of the invention

[0003] The present application provides a method for detecting deformation of a pin product by expanding the field of view, which is used to solve the technical problems of limited deformation detection range and insufficient detection accuracy in the prior art.

[0004] In view of the above problems, the present application provides a method for detecting pin product deformation by expanding the field of view.

[0005] The present application provides a method for detecting deformation of a pin product by expanding the field of view, the method comprising: Through the depth perception camera, the initial product image within the initial field of view of the target pin product to be inspected is collected; according to the grid division rule, the product image is divided to obtain multiple grid initial images in multiple grids, and deformation recognition is performed respectively to obtain multiple initial deformation information; when there is deformation in the multiple initial deformation information, the deformation trend of the multiple edge grid initial images is predicted according to the multiple edge initial deformation information of the multiple edge grids to obtain multiple deformation trend information; the field of view expansion coefficient is calculated and obtained by combining the multiple edge initial deformation information and the multiple deformation trend information, the initial field of view range is adjusted to obtain the expanded field of view range, the deformation detection of the target pin product is continued, and the final deformation detection result is obtained by combining the multiple initial deformation information.

[0006] One or more technical solutions provided in this application have at least the following technical effects or advantages: The present application uses a depth perception camera to collect an initial product image within an initial field of view of a target pin product to be inspected; divides the product image according to a grid division rule to obtain multiple grid initial images within multiple grids, performs deformation recognition on each of the multiple grids, and obtains multiple initial deformation information; when deformation exists in the multiple initial deformation information, predicts deformation trends of the multiple edge grid initial images based on multiple edge initial deformation information of multiple edge grids to obtain multiple deformation trend information; calculates a field of view expansion coefficient based on the multiple edge initial deformation information and multiple deformation trend information, adjusts the initial field of view to obtain an expanded field of view, continues to perform deformation detection of the target pin product, and obtains a final deformation detection result based on the multiple initial deformation information. The present invention solves the technical problems of limited deformation detection range and insufficient detection accuracy in the prior art. It collects the initial image of the target pin product through a depth perception camera, performs deformation recognition according to grid division, obtains initial deformation information, predicts the deformation trend according to the edge grid deformation information, calculates the field of view expansion coefficient, adjusts the field of view, and obtains more deformation information, thereby achieving the technical effect of improving the deformation detection accuracy and comprehensiveness of the target pin product. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0008] Figure 1 A schematic flow chart of a method for detecting pin product deformation by expanding the field of view provided in an embodiment of the present application; Figure 2 A schematic diagram of a process for obtaining multiple deformation trend information in a pin product deformation detection method by expanding the field of view provided in an embodiment of the present application. DETAILED DESCRIPTION

[0009] The present application provides a method for detecting deformation of a pin product by expanding the field of view, which is used to solve the technical problems of limited deformation detection range and insufficient detection accuracy in the prior art. The initial image of the target pin product is collected by a depth perception camera, deformation recognition is performed according to grid division, initial deformation information is obtained, deformation trend is predicted according to edge grid deformation information, the field of view expansion coefficient is calculated, the field of view range is adjusted, and more deformation information is obtained, thereby achieving the technical effect of improving the accuracy and comprehensiveness of deformation detection of the target pin product.

[0010] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of this application.

[0011] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or modules that are not explicitly listed or inherent to these processes, methods, products or devices.

[0012] Examples, such as Figure 1 As shown, the present application provides a method for detecting pin product deformation by expanding the field of view, the method comprising: Step S100: using a depth perception camera, an initial product image within an initial field of view of a target pin product to be inspected is collected.

[0013] Furthermore, the method provided in the application embodiment also includes: The depth perception camera is a 2.5D camera.

[0014] In an embodiment of the present application, first, an image of the target pin product to be inspected is captured by a depth perception camera. The depth perception camera is a 2.5D camera that can simultaneously obtain a two-dimensional image and depth information of the target product. The 2.5D camera uses its built-in depth perception technology to capture images containing depth data for each pixel, thereby providing richer spatial information. The initial field of view of the depth perception camera is determined by parameters such as focal length and viewing angle that are pre-set by humans. The depth perception camera is used to capture images of the target pin product to be inspected, and an initial product image within the initial field of view is obtained.

[0015] Step S200: Divide the product image according to a grid division rule to obtain a plurality of grid initial images in a plurality of grids, perform deformation recognition on each of the images, and obtain a plurality of initial deformation information.

[0016] In the embodiment of the present application, when dividing the product image, the product image is first divided into multiple grids of the same size according to the grid division rule, including edge grids and central grids. Each grid contains a part of the image area, and the grid size is preset.

[0017] Subsequently, deformation recognition is performed on the initial images in these grids to determine whether each grid is deformed, and the initial deformation information of each grid is obtained. The initial deformation information is the ratio of the deformation size to the product size.

[0018] Furthermore, in the method provided in the embodiment of the application, the product image is divided according to the grid division rule to obtain multiple grid initial images in multiple grids, and deformation recognition is performed respectively to obtain multiple initial deformation information, and further includes: The product image is divided according to a grid division rule to obtain a plurality of grid initial images within the plurality of grids, wherein the grid division rule includes a plurality of grids of the same size, and the plurality of grids include edge grids and a central grid; deformation recognition is performed on the plurality of grid initial images to obtain a plurality of initial deformation information.

[0019] In an embodiment of the present application, when performing deformation detection on a target pin product, the collected initial product image is first processed. According to the grid division rule, the image is divided into multiple grid areas of the same size. This grid size is pre-set to ensure that each grid can effectively cover the key information in the image while maintaining appropriate computational complexity. The grid division includes edge grids and central grids, wherein the edge grids are located around the image and may contain interference factors; while the central grid is located in the middle area of ​​the image, contains more detailed information, and is the key area for deformation recognition.

[0020] After the grid division is completed, the initial image in each grid is deformed. Specifically, first, based on the historical deformation detection data of the pin product, the sample initial grid image set is collected and processed, and the deformation size ratio in each sample grid image is marked to generate the sample initial deformation information set. Then, these sample initial images and deformation information with deformation information marked are used as supervised training data, and trained using a convolutional neural network (CNN) to learn the relationship between the image and the deformation. After the training is completed, multiple grid initial images are input into the trained deformation recognition path, and the initial deformation information of each grid is output after recognition.

[0021] Furthermore, in the method provided in the embodiment of the application, deformation recognition is performed on the multiple grid initial images to obtain multiple initial deformation information, and further includes: According to the historical deformation detection data of the pin product, a set of sample initial raster images is collected and processed, and the size ratio of the pin product deformation in each sample initial raster image is marked to obtain a set of sample initial deformation information; the sample initial raster image set and the sample initial deformation information set are used as supervised training data, and a deformation recognition path is trained based on a convolutional neural network; the multiple raster initial images are input into the deformation recognition path, and multiple initial deformation information are obtained by recognition output.

[0022] In the embodiment of the present application, the historical deformation detection data of the pin product is first obtained from the historical database, and then the product image in the historical deformation detection data is divided into multiple small blocks according to the grid division rule, and each small block is a grid image. Through this process, a sample initial grid image set is obtained.

[0023] Next, the size ratio of the pin product deformation in each sample initial grid image is marked. This process is carried out by manual annotation, and the annotation process includes calculating the ratio of the deformation size to the original size of the product. For example, if a pin product in the grid has a bulge, and the height of the bulge is 0.1mm, and the thickness of the pin is 1mm, then the size ratio of the deformation is 10%. Through this process, the initial deformation information set of the sample is obtained.

[0024] Subsequently, the sample initial raster image set and the sample initial deformation information set are used as supervised training data and trained through a convolutional neural network. A convolutional neural network is a deep learning algorithm that excels at automatically extracting features from images and performing classification or regression tasks. During the training process, the network continuously adjusts parameters to learn how to identify the deformation patterns and features of the pin products from the image. After training, CNN can automatically judge and output the corresponding deformation information based on the input raster image. After training, the deformation recognition path is obtained.

[0025] Finally, multiple grid initial images are input into the trained deformation recognition path, and the model analyzes the learned features and outputs the initial deformation information of each grid. Multiple initial deformation information is obtained through this process.

[0026] Step S300: when deformation exists in the plurality of initial deformation information, deformation trend prediction is performed on the plurality of edge grid initial images according to the plurality of edge initial deformation information of the plurality of edge grids to obtain a plurality of deformation trend information.

[0027] In an embodiment of the present application, when deformation exists in multiple initial deformation information, deformation information of multiple edge grids and corresponding initial images of edge grids are first screened out. Next, a deformation trend prediction path is constructed based on an integrated learning method, which includes multiple deformation prediction branches. After that, the ratio of each edge initial deformation information to the maximum value of all edge initial deformation information is calculated, and the ratio is multiplied by the total number of deformation prediction branches and rounded to obtain the number of multiple deformation prediction branches. Then, multiple edge grid initial images are input into the corresponding number of deformation prediction branches for deformation trend prediction. Finally, the final multiple deformation trend information is obtained by calculating the mean of multiple predicted deformation trend information.

[0028] Further, such as Figure 2 As shown, in the method provided by the embodiment of the application, when the multiple initial deformation information is deformed, a deformation trend prediction is performed on the multiple edge grid initial images according to the multiple edge initial deformation information of the multiple edge grids to obtain multiple deformation trend information, and further includes: When any one of the multiple initial deformation information is greater than 0, multiple edge initial deformation information of multiple edge grids and multiple edge grid initial images are screened; based on ensemble learning, a deformation trend prediction branch including the total number of deformation prediction branches is constructed to obtain a deformation trend prediction path; the ratio of each edge initial deformation information to the maximum value of the multiple edge initial deformation information is calculated respectively, and the ratios are multiplied by the total number of deformation prediction branches and rounded to obtain multiple numbers of deformation prediction branches; the multiple edge grid initial images are respectively input into the deformation trend prediction branches of the multiple numbers of deformation prediction branches, and multiple predicted deformation trend information sets are obtained by prediction output; the mean of the multiple predicted deformation trend information sets is calculated respectively to obtain multiple deformation trend information.

[0029] In the embodiment of the present application, when any one of the multiple initial deformation information is greater than 0, multiple edge initial deformation information of the multiple edge grids obtained by the aforementioned division and multiple edge grid initial images are obtained.

[0030] Next, based on ensemble learning, according to the deformation detection data of historical pin products, a set of sample edge grid initial images is collected, and a set of sample deformation trend information is identified. The deformation trend information includes the deformation probability of each sample, which is determined according to the proportion of deformation in the sample edge grid initial image away from the central grid area. Next, the sample image and deformation trend information are divided into multiple supervised training data sets to construct deformation trend prediction branches including the total number of deformation prediction branches. By training multiple deformation prediction branches separately based on an integrated convolutional neural network (CNN), a deformation trend prediction path is finally obtained. Among them, the total number of deformation prediction branches is pre-set.

[0031] Then, the ratio of each edge initial deformation information to the maximum value of multiple edge initial deformation information is calculated. This ratio reflects the severity of the deformation of each edge grid relative to other grids. Next, the ratio is multiplied by the total number of deformation prediction branches and rounded to an integer value, which represents the number of deformation prediction branches corresponding to each edge grid. Through this process, multiple deformation prediction branch numbers are obtained.

[0032] After the number of deformation prediction branches is obtained, multiple edge grid initial images are input into the corresponding number of deformation prediction branches for processing. Each branch will independently process the input image and output the prediction result. Among them, the deformation prediction branch is randomly selected in the deformation trend prediction path according to the number of deformation prediction branches. Through this process, multiple predicted deformation trend information sets are obtained.

[0033] Finally, multiple final deformation trend information is obtained by calculating the mean of multiple predicted deformation trend information sets respectively.

[0034] Furthermore, in the method provided in the embodiment of the application, based on ensemble learning, a deformation trend prediction branch including the total number of deformation prediction branches is constructed to obtain a deformation trend prediction path, and further includes: According to the deformation detection data of pin products in historical time, a set of sample edge grid initial images is collected, and according to the proportion of deformation in the area away from the central grid direction of different sample edge grid initial images, a set of sample deformation trend information is identified, wherein each sample deformation trend information includes a sample deformation probability; the sample edge grid initial image set and the sample deformation trend information set are divided to obtain multiple supervised training data of the total number of deformation prediction branches; using the multiple supervised training data, based on the integrated convolutional neural network, the deformation trend prediction branches of the total number of deformation prediction branches are trained respectively, and after the training is completed, a deformation trend prediction path is obtained.

[0035] In the embodiment of the present application, deformation detection data is first collected from the historical database to obtain a sample edge grid initial image set. The historical database contains image data collected during the past deformation detection process of the pin product, recording the deformation of the pin product in different time periods. By analyzing these historical data, a sample edge grid initial image set is obtained.

[0036] Next, according to the proportion of deformation in the area away from the central grid direction of different sample edge grid initial images, the sample deformation trend information set is identified. Among them, the area away from the central grid direction refers to the part outside the edge of the image, that is, the area not collected by the camera. This part is manually detected and the deformation information is artificially saved in the historical database. Different sample edge grid initial images refer to sample edge grid initial images with different initial deformation information. According to the proportion of deformation in the area away from the central grid direction of different sample edge grid initial images, the proportion obtained by calculating the number of deformation in the area away from the central grid direction of sample edge grid initial images with the same initial deformation information divided by the number of sample edge grid initial images with the same initial deformation information is used as the sample deformation trend information, that is, the sample deformation probability. Then, according to the calculated proportion, the sample edge grid initial image set is identified to obtain the sample deformation trend information set.

[0037] Then, the sample edge grid initial image set and the sample deformation trend information set are divided into multiple supervised training data according to the number of total deformation prediction branches. Each data includes the edge grid image and the corresponding deformation trend information, which is used to train the model.

[0038] Finally, multiple supervised training data are used to train deformation trend prediction branches of the total deformation prediction branch number based on the integrated convolutional neural network. Through this process, multiple deformation trend prediction branches of the total deformation prediction branch number are trained respectively, and the final deformation trend prediction path is obtained through the trained network model, which is used to predict the deformation trend of new samples.

[0039] Step S400: Combine the multiple edge initial deformation information and multiple deformation trend information to calculate the field of view expansion coefficient, adjust the initial field of view to obtain an expanded field of view, continue to perform deformation detection on the target pin product, and combine the multiple initial deformation information to obtain the final deformation detection result.

[0040] In the embodiment of the present application, firstly, by respectively calculating the ratio of each edge initial deformation information to the average of multiple edge initial deformation information, multiple deformation trend correction coefficients are obtained. Then, these correction coefficients are respectively applied to multiple deformation trend information to obtain multiple corrected deformation trend information. By calculating the average of these corrected deformation trend information, a field of view expansion coefficient is obtained. The field of view expansion coefficient is used to adjust the initial field of view range to obtain an expanded field of view range.

[0041] Then, the depth perception camera is used to collect the enlarged product image of the target pin product according to the enlarged field of view. The area of ​​the initial product image in the enlarged product image is cropped, and the area outside the initial product image in the enlarged product image is retained to perform deformation recognition and obtain the deformation information of the enlarged range. Finally, the deformation information of the enlarged range is combined with multiple initial deformation information to obtain the final deformation detection result.

[0042] Furthermore, in the method provided in the embodiment of the application, the field of view expansion coefficient is calculated by combining the multiple edge initial deformation information and the multiple deformation trend information, and the initial field of view range is adjusted to obtain the expanded field of view range, and the method also includes: Respectively calculate the ratio of each edge initial deformation information to the average of the multiple edge initial deformation information to obtain multiple deformation trend correction coefficients; respectively use the multiple deformation trend correction coefficients to multiply the multiple deformation trend information to obtain multiple corrected deformation trend information; calculate the average of the multiple corrected deformation trend information to obtain a field of view expansion coefficient; use the field of view expansion coefficient to multiply the initial field of view range to adjust and obtain an expanded field of view range.

[0043] In the embodiment of the present application, firstly, the ratio of each edge initial deformation information to the average of multiple edge initial deformation information is calculated to obtain multiple deformation trend correction coefficients.

[0044] Next, the obtained multiple deformation trend correction coefficients are used to correct the corresponding multiple deformation trend information. Specifically, each correction coefficient is multiplied by its corresponding deformation trend information to obtain multiple corrected deformation trend information.

[0045] After obtaining a plurality of corrected deformation trend information, the plurality of corrected deformation trend information are averaged and the calculation result is used as the field of view expansion coefficient.

[0046] The field of view expansion coefficient is then multiplied by the initial field of view to adjust the expanded field of view.

[0047] Furthermore, in the method provided in the embodiment of the application, the deformation detection of the target pin product is continued, and the final deformation detection result is obtained by combining the multiple initial deformation information, and further includes: Through the depth perception camera, an enlarged product image of the target pin product is collected according to the expanded field of view; the product image of the area outside the initial product image in the enlarged product image is cropped, deformation recognition is performed, and expanded range deformation information is obtained; and the expanded range deformation information and the multiple initial deformation information are combined to obtain the final deformation detection result.

[0048] In the embodiment of the present application, firstly, an enlarged product image of the target pin product is captured by a depth perception camera according to the enlarged field of view calculated above.

[0049] Next, the product image of the area outside the initial product image in the enlarged product image is cropped. Specifically, the enlarged product image is cropped to retain the area outside the initial product image. This means that the area covered by the original product image is removed from the enlarged product image, and the newly added area beyond the initial field of view is retained.

[0050] Then, deformation recognition is performed on the cropped additional area. Specifically, the cropped additional area is divided according to a grid division rule to obtain multiple grid images, which are input into the deformation recognition path obtained by the above training to obtain deformation information of the expanded range.

[0051] Finally, the expanded range deformation information obtained in the newly added area is combined with multiple initial deformation information within the initial field of view to obtain the final deformation detection result of the target pin product.

[0052] In the embodiments of the present application, in summary, the embodiments of the present application have at least the following technical effects: The present application uses a depth perception camera to collect an initial product image within an initial field of view of a target pin product to be inspected; divides the product image according to a grid division rule to obtain multiple grid initial images within multiple grids, performs deformation recognition on each of the multiple grids, and obtains multiple initial deformation information; when deformation exists in the multiple initial deformation information, predicts deformation trends of the multiple edge grid initial images based on multiple edge initial deformation information of multiple edge grids to obtain multiple deformation trend information; calculates a field of view expansion coefficient based on the multiple edge initial deformation information and multiple deformation trend information, adjusts the initial field of view to obtain an expanded field of view, continues to perform deformation detection of the target pin product, and obtains a final deformation detection result based on the multiple initial deformation information. The present invention solves the technical problems of limited deformation detection range and insufficient detection accuracy in the prior art. It collects the initial image of the target pin product through a depth perception camera, performs deformation recognition according to grid division, obtains initial deformation information, predicts the deformation trend according to the edge grid deformation information, calculates the field of view expansion coefficient, adjusts the field of view, and obtains more deformation information, thereby achieving the technical effect of improving the deformation detection accuracy and comprehensiveness of the target pin product.

[0053] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. The processes depicted in the accompanying drawings do not necessarily require the specific order and continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0054] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application shall be included in the protection scope of the present application.

[0055] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A method for detecting deformation of a pin product by expanding the field of view, characterized in that: The method comprises: Using a depth perception camera, an initial product image within an initial field of view of the target pin product to be inspected is collected; According to the grid division rule, the product image is divided to obtain a plurality of grid initial images in the plurality of grids, and deformation recognition is performed respectively to obtain a plurality of initial deformation information; When the multiple initial deformation information are deformed, a deformation trend prediction is performed on the multiple edge grid initial images according to the multiple edge initial deformation information of the multiple edge grids to obtain multiple deformation trend information; By combining the multiple edge initial deformation information and multiple deformation trend information, the field of view expansion coefficient is calculated, the initial field of view range is adjusted to obtain an expanded field of view range, and the deformation detection of the target pin product is continued. By combining the multiple initial deformation information, the final deformation detection result is obtained.

2. The method for detecting pin product deformation by expanding the field of view according to claim 1, characterized in that: The depth perception camera is a 2.5D camera.

3. The method for detecting pin product deformation by expanding the field of view according to claim 1, characterized in that: According to the grid division rule, the product image is divided to obtain a plurality of grid initial images in a plurality of grids, and deformation recognition is performed respectively to obtain a plurality of initial deformation information, including: Dividing the product image according to a grid division rule to obtain a plurality of grid initial images in a plurality of grids, wherein the grid division rule includes a plurality of grids of the same size, and the plurality of grids include edge grids and a central grid; Deformation recognition is performed on the multiple grid initial images to obtain multiple initial deformation information.

4. The method for detecting pin product deformation by expanding the field of view according to claim 3, characterized in that: Performing deformation recognition on the multiple grid initial images to obtain multiple initial deformation information includes: According to the historical deformation detection data of the pin product, a sample initial raster image set is collected and processed, and the size ratio of the pin product deformation in each sample initial raster image is marked to obtain a sample initial deformation information set; Using the sample initial grid image set and the sample initial deformation information set as supervised training data, and training a deformation recognition path based on a convolutional neural network; The multiple grid initial images are input into the deformation recognition path, and the recognition output is used to obtain multiple initial deformation information.

5. The method for detecting pin product deformation by expanding the field of view according to claim 1, characterized in that: When the multiple initial deformation information is deformed, deformation trend prediction is performed on the multiple edge grid initial images according to the multiple edge initial deformation information of the multiple edge grids to obtain multiple deformation trend information, including: When any one of the plurality of initial deformation information is greater than 0, screening a plurality of edge initial deformation information of a plurality of edge grids and a plurality of edge grid initial images; Based on ensemble learning, a deformation trend prediction branch including the total number of deformation prediction branches is constructed to obtain a deformation trend prediction path; Calculating the ratio of each edge initial deformation information to the maximum value of the multiple edge initial deformation information respectively, multiplying by the total number of deformation prediction branches respectively and rounding, to obtain multiple numbers of deformation prediction branches; Inputting the plurality of edge grid initial images into the deformation trend prediction branches of the plurality of deformation prediction branches respectively, and outputting the predictions to obtain a plurality of predicted deformation trend information sets; The average values ​​of the plurality of predicted deformation trend information sets are calculated respectively to obtain a plurality of deformation trend information.

6. The method for detecting pin product deformation by expanding the field of view according to claim 5, characterized in that: Based on ensemble learning, a deformation trend prediction branch including the total number of deformation prediction branches is constructed to obtain a deformation trend prediction path, including: According to the deformation detection data of the pin products in the historical time, a set of sample edge grid initial images is collected, and according to the proportion of deformation in the area away from the central grid direction of different sample edge grid initial images, a set of sample deformation trend information is obtained by marking, wherein each sample deformation trend information includes the sample deformation probability; Dividing the sample edge grid initial image set and the sample deformation trend information set to obtain multiple supervised training data of the total deformation prediction branch number; The plurality of supervised training data are used to train deformation trend prediction branches of the total number of deformation prediction branches based on an integrated convolutional neural network, and after the training is completed, a deformation trend prediction path is obtained.

7. The method for detecting pin product deformation by expanding the field of view according to claim 1, characterized in that: Combining the plurality of edge initial deformation information and the plurality of deformation trend information, calculating and obtaining a field of view expansion coefficient, and adjusting the initial field of view range to obtain an expanded field of view range, including: Respectively calculating the ratio of each edge initial deformation information to the average of the plurality of edge initial deformation information to obtain a plurality of deformation trend correction coefficients; The plurality of deformation trend correction coefficients are respectively used to multiply the plurality of deformation trend information to obtain a plurality of corrected deformation trend information; Calculating the average of the plurality of modified deformation trend information to obtain a field of view expansion coefficient; The field of view expansion coefficient is multiplied by the initial field of view to adjust and obtain the expanded field of view.

8. The method for detecting pin product deformation by expanding the field of view according to claim 1, characterized in that: Continuing to perform deformation detection of the target pin product, combining the multiple initial deformation information to obtain a final deformation detection result, including: By using a depth perception camera, an enlarged product image of the target pin product is collected according to the enlarged field of view; cropping the product image in the area outside the initial product image within the enlarged product image, performing deformation recognition, and obtaining deformation information of the enlarged range; The expanded range deformation information and the multiple initial deformation information are combined to obtain a final deformation detection result.

Images