Transmission full-automatic assembly detection method and system based on machine vision

Abstract

The invention discloses a full-automatic transmission assembly detection method and system based on machine vision, and the method comprises the steps: improving a scanning factor of a hiking algorithm through a spiral behavior process of a whale optimization algorithm, improving a snake vulture algorithm through a Levy flight process, and fusing the improved hiking algorithm into the hiking algorithm to jump out of local optimization to obtain a global optimal solution; according to the method, an improved combined intelligent algorithm is obtained, then the improved combined intelligent algorithm is used for optimizing a target detection model, a new target detection model is used for training, the type of the transmission is recognized, and the intelligent assembly detection recognition precision is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent detection, and in particular relates to a fully automatic transmission assembly detection method and system based on machine vision. Background Art

[0002] The statements herein merely provide background information related to the present invention and do not necessarily constitute prior art.

[0003] The transmission is a vital component of modern automobiles, and its assembly quality directly impacts key vehicle attributes such as reliability, safety, and comfort. As a key component, the transmission valve body is constructed from a variety of precision metal parts with a complex internal structure. Ball valves vary in diameter, color, and position, making them prone to missing, overinstalled, or incorrect assembly. Incorrect ball valve assembly can lead to oil channel blockage or leakage, impacting the transmission's hydraulic control and shifting smoothness, ultimately affecting the vehicle's proper function. Proper assembly and testing are crucial for ensuring quality.

[0004] As a detection method, machine vision can obtain the required useful information more quickly than other detection methods, and can automatically process the obtained information. It is also easy to integrate with design information and processing control information. However, the existing transmission assembly recognition method is affected by the different structural sizes of the transmission valve body and cannot guarantee high recognition accuracy. Summary of the Invention

[0005] In view of the shortcomings of the existing technology, the purpose of the present invention is to provide a fully automatic assembly detection method and system for transmissions based on machine vision. The present invention improves the scanning factor of the hiking algorithm by using the spiral behavior process of the whale optimization algorithm, uses the Levy flight process to improve the secretary vulture algorithm and then integrates it into the hiking algorithm to jump out of local optimization and obtain the global optimal solution, thereby obtaining an improved combined intelligent algorithm. The improved combined intelligent algorithm is then used to optimize the target detection model, and the new target detection model is used for training to identify the type of transmission, thereby greatly improving the recognition accuracy of intelligent assembly detection.

[0006] In order to achieve the above object, the present invention is implemented through the following technical solutions:

[0007] A fully automatic transmission assembly detection method based on machine vision includes the following steps:

[0008] Acquire image data of the transmission;

[0009] Using a combined intelligent algorithm to improve the target detection model, and training the improved target detection model based on the image data;

[0010] Using the trained target detection model to identify the image data of the target transmission, a recognition result is obtained;

[0011] Among them, the process of using the combined intelligent algorithm to improve the target detection model is as follows: the spiral behavior process of the whale optimization algorithm is used to improve the optimization process of the hiking algorithm, the scan factor is updated in combination with the position of the hiker and the position of the whale, and the updated scan factor is used to update the position of the hiker. The secretary vulture algorithm is improved by using the Levy flight process and then integrated into the hiking algorithm to jump out of the local optimization and obtain the global optimal solution. The optimization process is iterated until the iteration requirements are met, forming a combined intelligent algorithm to obtain the final optimal solution. The final optimal solution is used as the number and weight of the Bottleneck modules in the C2f module of the target detection model to achieve the improvement of the target detection model.

[0012] As an optional implementation, the process of acquiring transmission image data includes acquiring images of key assembly components of the transmission valve body, identifying the color and size of each key assembly component, and establishing corresponding data sets for different models of transmission valve bodies.

[0013] As an optional implementation, the spiral behavior process of the whale optimization algorithm is used to improve the optimization process of the hiking algorithm. The process of updating the scan factor based on the position of the hiker and the position of the whale includes:

[0014] The whale optimization algorithm searches in a spiral manner. The spiral behavior strategy is to generate a spiral path between the current position and the position of the global optimal solution. The whale moves along this path. The spiral behavior process of an individual whale is as follows:

[0015]

[0016] Where t is the number of iterations, b is a constant that defines the spiral hunting, l is a random number between [-1,1], and X i (t) represents the position of the individual whale at the tth iteration, is the individual with the best position in t iterations, is the position of the leading hiker in the hiking algorithm, that is, the optimal value of the hiking algorithm, and a is the parameter for adjusting the whale spiral distance, which is calculated as follows:

[0017]

[0018] During the walker's search, its speed is determined by the whale optimization algorithm's scan factor α i,t Improvements are made, specifically:

[0019]

[0020] During the search iteration, α i,t The value of decreases continuously, which makes hikers tend to explore and search locally.

[0021] As an optional implementation, the process of improving the secretary bird algorithm using the Levy flight process includes:

[0022] The Levy flight distribution function is:

[0023]

[0024] Where η is a constant with a value of 1.5, u and v are random numbers between [0, 1], and the calculation formula for σ is:

[0025]

[0026] Where, Γ represents the gamma function;

[0027] The escape strategy of the improved secretary bird algorithm is:

[0028]

[0029] Where r is a random number between [0,1], T is the total number of iterations, RB is a random number matrix generated from a standard normal distribution with a mean of 0 and a standard deviation of 1, R is an array of dimension (1×j) randomly generated from a normal distribution, K is a random choice of 1 or 2, and x is best is the optimal solution of the current iteration of the eagle algorithm, x random is the random solution of the current iteration, x i,j is the current position of the secretary bird. This is the new location of the secretary bird after it escaped.

[0030] As an optional implementation, the Levy flight process is used to improve the secretary vulture algorithm and then integrated into the walking tour algorithm to break away from the local optimization. The process of obtaining the global optimal solution includes the following steps:

[0031] make It represents the difference between the hiker's position after iteration t+1 and iteration t divided by the hiker's position at iteration t, β i,t is the current position of hiker i, β i,t+1 is the position of hiker i at the next moment;

[0032] When the number of consecutive set values ​​Z(t+1) is less than the set value during the optimization process, the position at this time is recorded as Use the escape strategy of the improved secretary bird algorithm to migrate the individual, and the new position is Recalculate the scan factor α using the new position i,t The value of , and then use the scanning factor α in the iterative process i,tPerform optimization.

[0033] As a further limited implementation method, when the number of consecutive set values ​​Z(t+1) is less than the set value during the optimization process, the new optimal solution is compared with the previous optimal solution, the best of the two is selected, and the number of jumps is recorded at the same time. When the number of jumps reaches the set value or the number of iterations reaches the maximum, the iteration stops and the final optimal position is obtained.

[0034] As an optional implementation, the process of using the final optimal solution as the number and weight of Bottleneck modules in the C2f module of the target detection model includes: the C2f module is composed of a Conv module, a Split module, a Bottleneck module, and a Concat module, wherein the Conv module is used to perform convolution calculation processing on the input feature map, and the Split module is used to split the feature map after the convolution layer processing into two parts, and the split feature maps still maintain the same spatial size, and the number of channels in each part is the same;

[0035] The Bottleneck module is used to perform two convolutional layer processes on the segmented feature map, and the Concat module merges the outputs of the Bottleneck module according to the channel dimension to form a spliced ​​feature map, and reduces it to the final output.

[0036] As a further implementation, the calculation process of the Bottleneck module is:

[0037] Bottleneck N =ω1·Conv 1,n +ω2·Conv 2,n ,N=3,...,15

[0038] Where N represents the number of Bottleneck modules, and the range of N is [3,15]. Its value is obtained by optimizing the combined intelligent algorithm. Conv1 and Conv2 are two Conv modules. ω1 and ω2 are weight coefficients. ω1+ω2=1, and the values ​​of ω1 and ω2 range from (0, 1). The value of ω1 or ω2 is optimized using the combined intelligent algorithm, and the value of the other weight coefficient is calculated according to ω1+ω2=1.

[0039] As an optional implementation, during the training of the improved target detection model, the fitness function is defined as mAP, and the calculation formula is:

[0040]

[0041] Where P is precision, R is recall, AP is average precision, mAP is mean average precision, TP is the number of correctly detected positive samples, FP is the number of negative samples incorrectly detected as positive samples, FN is the number of positive samples incorrectly detected as negative samples, k is the number of categories, and the maximum fitness function is the training objective function.

[0042] A fully automatic transmission assembly and detection system based on machine vision, comprising:

[0043] an image acquisition module configured to acquire image data of the transmission;

[0044] a combined algorithm improvement module, configured to improve the target detection model using a combined intelligent algorithm, and train the improved target detection model based on the image data;

[0045] a transmission recognition module configured to recognize the image data of the target transmission using the trained target detection model to obtain a recognition result;

[0046] Among them, the process of using the combined intelligent algorithm to improve the target detection model is as follows: the spiral behavior process of the whale optimization algorithm is used to improve the optimization process of the hiking algorithm, the scan factor is updated in combination with the position of the hiker and the position of the whale, and the updated scan factor is used to update the position of the hiker. The secretary vulture algorithm is improved by using the Levy flight process and then integrated into the hiking algorithm to jump out of the local optimization and obtain the global optimal solution. The optimization process is iterated until the iteration requirements are met, forming a combined intelligent algorithm to obtain the final optimal solution. The final optimal solution is used as the number and weight of the Bottleneck modules in the C2f module of the target detection model to achieve the improvement of the target detection model.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] The present invention provides a fully automatic transmission assembly detection method based on machine vision, and proposes to improve the scanning factor α of the walking tour algorithm by using the spiral behavior process of the whale optimization algorithm. i,t The Levy flight process is used to improve the secretary vulture algorithm and then integrated into the hiking algorithm to jump out of the local optimization and obtain the global optimal solution, so that the hiking algorithm can find the optimal point locally and jump out of the optimal point to a certain extent, realizing the algorithm's ability to jump out of the local optimal point in complex, multi-peak data problems.

[0049] The present invention combines intelligent algorithms to optimize the number and weight of the Bottleneck modules in the improved target detection model, and uses the new target detection model to train the target detection model to identify the type of transmission, which greatly improves the detection and recognition accuracy of the transmission fully automatic assembly robot.

[0050] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0052] Figure 1 This is a flow chart of a fully automatic transmission assembly and detection method based on machine vision in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of the present invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which the present invention belongs.

[0054] It should be noted that the terms used herein are only for describing specific embodiments and are not intended to limit the exemplary embodiments according to the present invention. As used herein, unless the context clearly indicates otherwise, the singular form is intended to include the plural form. In addition, it should be understood that when the terms "comprise" and / or "include" are used in this specification, they indicate the presence of features, steps, operations, devices, components and / or combinations thereof.

[0055] In the absence of conflict, the embodiments and features in the embodiments of this application can be combined with each other.

[0056] Example 1

[0057] As mentioned in the background technology, the existing transmission assembly recognition method is affected by the different structural sizes of the transmission valve body and cannot guarantee high recognition accuracy. In the transmission assembly recognition method of the present invention, the spiral behavior process of the whale optimization algorithm is first used to improve the scanning factor of the hiking algorithm, and the secretary vulture algorithm after the Levy flight process is improved is integrated into the hiking algorithm to jump out of the local optimization and obtain the global optimal solution, thereby obtaining a combined intelligent algorithm. Then, the combined intelligent algorithm is used to optimize the number and weight of the Bottleneck module in the C2f module of the target detection model, and the new target detection model is used to train the target detection model to identify the type of transmission, which greatly improves the intelligent assembly detection recognition accuracy.

[0058] like Figure 1As shown, this embodiment provides a fully automatic transmission assembly detection method based on machine vision, comprising the following steps:

[0059] Step 1: Acquire two-dimensional image data of the transmission to be inspected;

[0060] An industrial camera is used to capture images of the transmission valve body assembly, including but not limited to: valve blocks, sliders, large and small marbles, and large and small springs. A visual inspection system is used to identify the color and size of each valve body component. For different types of valve bodies, pre-processing is performed to establish a valve body visual inspection dataset including ball valves.

[0061] Step 2: Use the spiral behavior process of the whale optimization algorithm and the improved secretary vulture algorithm using the Levy flight process and integrate them into the hiking algorithm. Specifically:

[0062] The basic content of the hiking algorithm is the location update of the hiker in the hiking algorithm, specifically:

[0063] The initial slope formula of the hiking algorithm is:

[0064]

[0065] Where dh and dx represent the hiker’s height difference and travel distance difference, respectively, θ i,t is the slope angle of the path or terrain, S i,t represents the slope of the path or terrain. The initial speed of the hiker is given by the Tobler hiking function:

[0066]

[0067] Where W i,t It represents the speed of the initial hiker in the iteration time t, in km / h.

[0068] Initialize the population and use random numbers to randomly generate the initial positions of multiple hikers in the space. For the i-th hiker, its initial position is expressed as:

[0069] β i,0 =φ L +λ·(φ U -φ L )i=0,1,2,...,N

[0070] Where N represents the number of individuals, φ L is the lower bound of the space, φ U is the upper bound of the space, and λ is a randomly initialized number between 0 and 1.

[0071] The slope angle is initialized using the probability density function of the Weibull distribution:

[0072]

[0073] Here, x is a random variable, which is the initial slope angle faced by the hiker and is a random number between 0° and 50°. m>0 is the scale parameter, and η>0 is the shape parameter, which determines the shape of the distribution.

[0074] During the search process, the speed of the hiker is determined by the initial speed, the position of the leading hiker, the actual position of the hiker, and the scanning factor. Therefore, the current speed of the i-th hiker is:

[0075] W i,t =W i,t-1 +γ i,t (β best -α i,t β i,t )

[0076] Among them, γ i,t is a random number uniformly distributed between 0 and 1, α i,t is the scanning factor (SF) of hiker i, which ranges from [1,3] and ensures that hiker i does not stray too far from the leading hiker so that he can see the direction of the leading hiker and receive the signal from the leading hiker. i,t is the current position of hiker i, β best It is the position of the lead hiker, W i,t represents the current speed of hiker i, W i,t-1 represents the speed of hiker i at the last moment. According to the hiker’s speed, the new position update of hiker i should be:

[0077] β i,t+1 =β i,t +W i,t

[0078] Where, β i,t+1 is the updated position of hiker i. Hikers will continuously update their positions to reach the leader's position.

[0079] The whale optimization algorithm searches in a spiral manner. The spiral behavior strategy is to generate a spiral path between the current position and the position of the global optimal solution, and then the whale moves along this path. Specifically:

[0080] The spiral behavior process of an individual whale is:

[0081]

[0082] Where t is the number of iterations, b is a constant that defines the spiral hunting, l is a random number between [-1,1], and X iIndicates the current location of the individual whale. is the best-positioned individual, is the position of the leading hiker in the hiking algorithm, that is, the optimal value of the hiking algorithm, and a is the parameter for adjusting the whale spiral distance, which is calculated as follows:

[0083]

[0084] During the walker's search, its speed is determined by the whale optimization algorithm's scan factor α i,t Improvements are made, specifically:

[0085]

[0086] During the search iteration, α i,t The value of will continue to decrease, making hikers tend to explore and search locally.

[0087] The process of improving the Secretary Vulture Algorithm by using the Levy flight process and integrating it into the hiking algorithm to break away from local optimization and obtain the global optimal solution includes:

[0088] The Levy flight distribution function is:

[0089]

[0090] Where η is a constant with a value of 1.5, u and v are random numbers between [0, 1], and the calculation formula for σ is:

[0091]

[0092] Where Γ represents the gamma function, and η is a constant with a value of 1.5.

[0093] The Secretary Bird's escape strategy is:

[0094]

[0095] Where r is a random number between [0,1], T is the total number of iterations, RB is a random number matrix generated from a standard normal distribution with a mean of 0 and a standard deviation of 1, R is an array of dimension (1×j) randomly generated from a normal distribution, K is a random choice of 1 or 2, and x is random is the random solution of the current iteration, x i,j is the current position of the secretary bird. This is the new location of the Secretary Bird after it escaped. best is the optimal solution of the current iteration of the Secretary Vulture algorithm.

[0096] make It represents the difference between the hiker's position after iteration t+1 and iteration t divided by the hiker's position at iteration t, β i,tis the current position of hiker i, β i,t+1 is the location of hiker i at the next moment.

[0097] When the hiking combination algorithm continuously T n When Z(t+1) is less than the set value for the first time, it means that the hiker's position has changed very little, and the position at this time is recorded as The improved secretary bird escape strategy is used to relocate the individual to the new location. Recalculate the scan factor α using the new position i,j The value of T n When Z(t+1)<1% occurs for the first time, T n When the number of times is 25 to 40, compare the new optimal solution with the previous optimal solution, select the best one, and use M n Record the number of times. When the number of bounces reaches T nn times, T nn The iteration stops when the number of iterations reaches 5 to 10 or the maximum number of iterations is reached, and the optimal position is obtained. This resulted in a combined intelligent algorithm.

[0098] Step 3: Improve the Bottleneck module in the C2f module in the YOLOv8 algorithm, specifically:

[0099] The YOLOv8 algorithm architecture consists of a backbone and a head. The backbone is used for feature extraction and primarily includes the Conv module, the C2f module, and the SPPF module. The Conv module performs convolution operations, the C2f module performs feature extraction, and the SPPF module provides adaptively sized output.

[0100] The C2f module structure consists of a Conv module, a Split module, a Bottleneck module, and a Concat module.

[0101] The top is a Conv module, and the convolution calculation formula is:

[0102]

[0103] Where k is the convolution kernel size, s is the stride, p is whether to fill (if p = 1, it means to fill 1 pixel around the input; if p = 0, it means no filling), H out and W out is the height and width of the output feature map, H in and W in are the height and width of the input feature map.

[0104] The feature map after convolutional layer processing will be split into two parts, each with c channels. out , with a height of h and a width of w, the feature map after segmentation still maintains the same spatial size h×w, but the number of channels in each part is 0.5×c out .

[0105] Each segmented part will enter multiple Bottleneck modules respectively. In the Bottleneck module, the input feature map will be processed by two convolutional layers, and at the end, the input will be directly added to the output of the convolutional layer through a shortcut, so that a residual is formed between the input and output. The calculation formula is:

[0106] Bottleneck N =ω1·Conv 1,n +ω2·Conv 2,n ,N=3,...,15

[0107] Where N represents the number of Bottleneck modules, and the range of N is [3,15]. Its value is optimized by the combined intelligent algorithm, that is, it is given by the combined intelligent algorithm. Conv1 and Conv2 are two Conv modules, ω1 and ω2 are weight coefficients, ω1+ω2=1, and the value range of ω1 and ω2 is (0, 1). The value of ω1 or ω2 is optimized by the combined intelligent algorithm, that is, it is given by the combined intelligent algorithm, and the other value is given by calculation.

[0108] After all the Bottleneck modules are processed, all the feature maps will be merged by the Concat module according to the channel dimension. The size of the concatenated feature map is: h×w×(0.5×N+2)×c out .

[0109] The concatenated feature map is reduced to the final output through a convolutional layer. The output feature map size is: h×w×c out , for downstream network processing or target detection.

[0110] When using the YOLOv8 algorithm to train photos of assembly line tests, its fitness function is defined as mAP, and the calculation formula is:

[0111]

[0112] Where P is precision, R is recall, AP is average precision, mAP is mean average precision, TP is the number of correctly detected positive samples, FP is the number of negative samples incorrectly detected as positive samples, FN is the number of positive samples incorrectly detected as negative samples, k is the number of categories, and d() is the differential function.

[0113] The maximum fitness function is used as the objective function for the combined intelligent algorithm to find the optimal solution.

[0114] Step 4: Based on a variety of transmission image data of different sizes and types, use the improved YOLOv8 algorithm to perform transmission training and recognition.

[0115] The specific process of this embodiment can also be described as:

[0116] For the formed transmission valve body visual inspection dataset, mAP is used as its fitness function, and the obtained combined intelligent algorithm is used to optimize the parameters ω1 and ω2 in Bottleneck-N and the merged module, thereby obtaining the optimal transmission recognition model.

[0117] Example 2

[0118] A fully automatic transmission assembly and detection system based on machine vision, comprising:

[0119] an image acquisition module configured to acquire image data of the transmission;

[0120] a combined algorithm improvement module, configured to improve the target detection model using a combined intelligent algorithm, and train the improved target detection model based on the image data;

[0121] a transmission recognition module configured to recognize the image data of the target transmission using the trained target detection model to obtain a recognition result;

[0122] Among them, the process of using the combined intelligent algorithm to improve the target detection model is as follows: the spiral behavior process of the whale optimization algorithm is used to improve the optimization process of the hiking algorithm, the scan factor is updated in combination with the position of the hiker and the position of the whale, and the updated scan factor is used to update the position of the hiker. The secretary vulture algorithm is improved by using the Levy flight process and then integrated into the hiking algorithm to jump out of the local optimization and obtain the global optimal solution. The optimization process is iterated until the iteration requirements are met, forming a combined intelligent algorithm to obtain the final optimal solution. The final optimal solution is used as the number and weight of the Bottleneck modules in the C2f module of the target detection model to achieve the improvement of the target detection model.

[0123] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0124] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts 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, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0125] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0126] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0127] The foregoing description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations of the present invention are possible. Any modifications, equivalent substitutions, or improvements made by those skilled in the art that fall within the spirit and principles of the present invention and do not require creative effort are intended to be within the scope of protection of the present invention.

Claims

1. A fully automatic transmission assembly detection method based on machine vision, characterized in that: The steps include: Acquire image data of the transmission; Using a combined intelligent algorithm to improve the target detection model, and training the improved target detection model based on the image data; Using the trained target detection model to identify the image data of the target transmission, a recognition result is obtained; Among them, the process of using the combined intelligent algorithm to improve the target detection model is as follows: the spiral behavior process of the whale optimization algorithm is used to improve the optimization process of the hiking algorithm, the scan factor is updated in combination with the position of the hiker and the position of the whale, and the updated scan factor is used to update the position of the hiker. The secretary vulture algorithm is improved by using the Levy flight process and then integrated into the hiking algorithm to jump out of the local optimization and obtain the global optimal solution. The optimization process is iterated until the iteration requirements are met, forming a combined intelligent algorithm to obtain the final optimal solution. The final optimal solution is used as the number and weight of the Bottleneck modules in the C2f module of the target detection model to achieve the improvement of the target detection model.

2. The fully automatic transmission assembly and detection method based on machine vision according to claim 1, characterized in that: The process of acquiring transmission image data includes acquiring images of key assembly parts of the transmission valve body, identifying the colors and sizes of the key assembly parts, and establishing corresponding data sets for different models of transmission valve bodies.

3. The fully automatic transmission assembly and detection method based on machine vision according to claim 1, characterized in that: The spiral behavior process of the whale optimization algorithm is used to improve the optimization process of the hiking algorithm. The process of updating the scanning factor based on the position of the hiker and the position of the whale includes: The whale optimization algorithm searches in a spiral manner. The spiral behavior strategy is to generate a spiral path between the current position and the position of the global optimal solution. The whale moves along this path. The spiral behavior process of an individual whale is as follows: Where t is the number of iterations, b is a constant that defines the spiral hunting, l is a random number between [-1,1], and X i (t) represents the position of the individual whale at t iterations, is the individual with the best position in t iterations, is the position of the leading hiker in the hiking algorithm, that is, the optimal value of the hiking algorithm, and a is the parameter for adjusting the whale spiral distance, which is calculated as follows: During the walker's search, its speed is determined by the whale optimization algorithm's scan factor α i,t Improvements are made, specifically: During the search iteration, α i,t The value of decreases continuously, which makes hikers tend to explore and search locally.

4. The fully automatic transmission assembly and detection method based on machine vision according to claim 1, characterized in that: The process of improving the secretary bird algorithm using the Levy flight process includes: The Levy flight distribution function is: Where η is a constant with a value of 1.5, u and v are random numbers between [0, 1], and the calculation formula for σ is: Where, Γ represents the gamma function; The escape strategy of the improved secretary bird algorithm is: Where r is a random number between [0,1], T is the total number of iterations, RB is a random number matrix generated from a standard normal distribution with a mean of 0 and a standard deviation of 1, R is an array of dimension (1×j) randomly generated from a normal distribution, K is a random choice of 1 or 2, and x is best is the optimal solution of the current iteration of the eagle algorithm, x random is the random solution of the current iteration, x i,j is the current position of the secretary bird. This is the new location of the secretary bird after it escaped.

5. The fully automatic transmission assembly and detection method based on machine vision according to claim 1 or 4, characterized in that: The process of improving the Secretary Vulture Algorithm by using the Levy flight process and integrating it into the hiking algorithm to break away from local optimization and obtain the global optimal solution includes: make It represents the difference between the hiker's position after iteration t+1 and iteration t divided by the hiker's position at iteration t, β i,t is the current position of hiker i, β i,t+1 is the position of hiker i at the next moment; When the number of consecutive set values ​​Z(t+1) is less than the set value during the optimization process, the position at this time is recorded as Use the escape strategy of the improved secretary bird algorithm to migrate the individual, and the new position is Recalculate the scan factor α using the new position i,t The value of , and then use the scanning factor α in the iterative process i,t Perform optimization.

6. The fully automatic transmission assembly and detection method based on machine vision as claimed in claim 5, characterized in that: When the number of consecutive set values ​​Z(t+1) is less than the set value during the optimization process, the new optimal solution is compared with the previous optimal solution, and the best one is selected. At the same time, the number of jumps is recorded. When the number of jumps reaches the set value or the number of iterations reaches the maximum, the iteration stops and the final optimal position is obtained.

7. The fully automatic transmission assembly and detection method based on machine vision as claimed in claim 1, characterized in that: The process of using the final optimal solution as the number and weight of Bottleneck modules in the C2f module of the target detection model includes: the C2f module is composed of a Conv module, a Split module, a Bottleneck module, and a Concat module, wherein the Conv module is used to perform convolution calculation processing on the input feature map, and the Split module is used to split the feature map processed by the convolution layer into two parts, and the split feature maps still maintain the same spatial size and the number of channels of each part is the same; The Bottleneck module is used to perform two convolutional layer processes on the segmented feature map, and the Concat module merges the outputs of the Bottleneck module according to the channel dimension to form a spliced ​​feature map, and reduces it to the final output.

8. The fully automatic transmission assembly and detection method based on machine vision as claimed in claim 7, characterized in that: The calculation process of the Bottleneck module is: Bottleneck N =ω1·Conv 1,n +ω2·Conv 2,n ,N=3,...,15 Where N represents the number of Bottleneck modules, and the range of N is [3,15]. Its value is obtained by optimizing the combined intelligent algorithm. Conv1 and Conv2 are two Conv modules. ω1 and ω2 are weight coefficients. ω1+ω2=1, and the values ​​of ω1 and ω2 range from (0, 1). The value of ω1 or ω2 is optimized using the combined intelligent algorithm, and the value of the other weight coefficient is calculated according to ω1+ω2=1.

9. The fully automatic transmission assembly and detection method based on machine vision according to claim 1, characterized in that: During the training of the improved target detection model, the fitness function is defined as mAP, and the calculation formula is: Where P is precision, R is recall, AP is average precision, mAP is mean average precision, TP is the number of correctly detected positive samples, FP is the number of negative samples incorrectly detected as positive samples, FN is the number of positive samples incorrectly detected as negative samples, k is the number of categories, and the maximum fitness function is the training objective function.

10. A fully automatic transmission assembly and detection system based on machine vision, characterized in that: include: an image acquisition module configured to acquire image data of the transmission; a combined algorithm improvement module, configured to improve the target detection model using a combined intelligent algorithm, and train the improved target detection model based on the image data; a transmission recognition module configured to recognize the image data of the target transmission using the trained target detection model to obtain a recognition result; Among them, the process of using the combined intelligent algorithm to improve the target detection model is as follows: the spiral behavior process of the whale optimization algorithm is used to improve the optimization process of the hiking algorithm, the scan factor is updated in combination with the position of the hiker and the position of the whale, and the updated scan factor is used to update the position of the hiker. The secretary vulture algorithm is improved by using the Levy flight process and then integrated into the hiking algorithm to jump out of the local optimization and obtain the global optimal solution. The optimization process is iterated until the iteration requirements are met, forming a combined intelligent algorithm to obtain the final optimal solution. The final optimal solution is used as the number and weight of the Bottleneck modules in the C2f module of the target detection model to achieve the improvement of the target detection model.

Images