Automobile brake disc defect detection method and system

Abstract

The invention relates to the technical field of brake disc detection, in particular to an automobile brake disc defect detection method and system, and the method comprises the steps: obtaining the number and surface image of an automobile brake disc, and carrying out the image enhancement preprocessing of the surface image based on an improved Retinex algorithm; detecting and identifying the preprocessed surface image according to a preset improved AlexNet network model, and determining defect information on the automobile brake disc; marking the automobile brake disc according to the defect types and the number of the defect types so as to judge the quality of the automobile brake disc; according to the number of the automobile brake disc, the defect information and the mark of the automobile brake disc, determining whether the production line needs to be stopped for maintenance. According to the method, the overall contrast ratio of the surface image is improved, so that the local features of the surface image are more obvious, the improved AlexNet network model is adopted, so that the defect detection of the surface image can be more efficient and accurate, the production efficiency is improved, and the production cost is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of brake disc detection, and particularly to a method and system for detecting defects of automotive brake discs. Background Art

[0002] As an important component of a vehicle braking system, a brake disc is closely related to driving safety; defects such as pores and cracks generated during the casting of a brake disc will affect the lifespan of the brake disc. Therefore, the detection of brake discs is particularly important.

[0003] The existing detection technologies for brake discs are not yet mature enough, and there are situations where detection is carried out purely manually or in combination with machines, with insufficient automation, high time costs, low production efficiency, and relatively large errors, which cannot meet the requirements of enterprises for cost reduction and efficiency improvement. Summary of the Invention

[0004] To solve the above technical problems, the present invention provides a method and system for detecting defects of automotive brake discs.

[0005] The present invention adopts the following technical solutions: A method for detecting defects of automotive brake discs, the method comprising: Obtaining the number and surface image of an automotive brake disc, and performing image enhancement preprocessing on the surface image based on an improved Retinex algorithm; Performing detection and recognition on the preprocessed surface image according to a preset improved AlexNet network model to determine the defect information on the automotive brake disc; wherein, the defect information includes the defect type and the quantity corresponding to the defect type; Marking the automotive brake disc according to the defect type and the quantity of the defect type to judge the quality of the automotive brake disc; Confirming whether the production line needs to stop production for maintenance according to the number of the automotive brake disc, the defect information, and the marking of the automotive brake disc.

[0006] The method for detecting defects of an automotive brake disc according to an embodiment of the present invention enhances the surface image by using an improved Retinex algorithm, effectively improving the overall contrast of the surface image, making the local features of the automotive brake disc more obvious, and providing a basic guarantee for improving the detection and recognition of the AlexNet network model; by using the improved AlexNet network model, the learning ability of the model is improved, the model recognition speed is accelerated, and the defect detection of the surface image can achieve an efficient and accurate effect; through different defect types and the number of defect types, the automotive brake disc is analyzed and judged at multiple levels to mark the automotive brake discs that do not meet the production quality. Finally, according to the marking results and the defect information of the automotive brake disc, it is judged whether there are defects in the production line. If so, the production line is shut down for maintenance to avoid greater losses and effectively reduce the production cost.

[0007] Further, the steps of performing image enhancement preprocessing on the surface image based on the improved Retinex algorithm specifically include: Replacing the logarithmic transformation in the Retinex algorithm with an adaptive power transformation to obtain an initial enhanced image:

[0008] Wherein, is the initial enhanced image, is the original image, is the Gaussian low-pass filter, is the adaptive power exponent, , is the information entropy of the surface image, are the row and column of the image matrix respectively; Using a sliding window of size to traverse the initial enhanced image , obtaining the sliding window average value and the sliding window variance, and obtaining the target enhanced image according to the sliding window average value and the sliding window variance, thereby completing the image enhancement preprocessing of the surface image; Wherein, the sliding window average value, the sliding window variance and the target enhanced image are:

[0009]

[0010]

[0011] Wherein, is the sliding window average value, is the sliding window variance, is the target enhanced image.

[0012] Furthermore, the improved AlexNet network model includes 6 convolutional layers, 4 pooling layers and 3 fully connected layers, and uses a BN layer to replace the LRN normalization layer in the traditional AlexNet network model; Among them, the convolution kernel size of the Conv1 convolutional layer is 13×13, the convolution kernel size of the Conv2 convolutional layer is 5×5, the convolution kernel sizes of the Conv3 convolutional layer, the Conv4 convolutional layer and the Conv5 convolutional layer are all 3×3, and the convolution kernel size of the Conv6 convolutional layer is 1×1.

[0013] Furthermore, the steps of marking the automotive brake disc according to the defect type and the number of the defect types to judge the quality of the automotive brake disc specifically include: The defects of the automotive brake disc are divided into hole defects and line defects. Denote the automotive brake disc detected and recognized by the improved AlexNet network model as , the number of the hole defects is denoted as , the number of the line defects is denoted as , where is the number of the automotive brake disc; If , then mark the automotive brake disc with the number as a defective product; where is the defect quantity threshold; Calibrate the defects of the automotive brake disc based on the pixel equivalent of shape matching to obtain the sizes of the defects of the automotive brake disc, which are respectively denoted as and , where is the size of the th hole defect of the automotive brake disc with the number , is the size of the th line defect of the automotive brake disc with the number ; If or any one of the conditions is satisfied, then mark the automotive brake disc with the number as a defective product; where is the first size threshold of the hole defect, is the first size threshold of the line defect; If or any one of the conditions is satisfied, then mark the automotive brake disc with the number as a defective product; where is the second size threshold for the hole defect, is the second size threshold for the line defect.

[0014] Further, the steps of confirming whether the production line needs to be shut down for maintenance according to the serial number of the vehicle brake disc, the defect information and the mark of the vehicle brake disc specifically include: Record the serial number as of the quality mark of the vehicle brake disc as , where when, the vehicle brake disc with the serial number is a defective product, when, the vehicle brake disc with the serial number is a qualified product; If , it is determined that the production line needs to be shut down for maintenance; If , and , meanwhile at least one of them is 0, it is determined that the production line needs to be shut down for maintenance; where is the number of hole defects of the vehicle brake disc with the serial number , is the number of hole defects of the vehicle brake disc with the serial number , is the number of line defects of the vehicle brake disc with the serial number , is the number of line defects of the vehicle brake disc with the serial number .

[0015] The present invention also proposes a vehicle brake disc defect detection system, and the system includes: An acquisition module, configured to acquire the serial number and the surface image of the vehicle brake disc, and perform image enhancement preprocessing on the surface image based on an improved Retinex algorithm; A detection module, configured to perform detection and recognition on the preprocessed surface image according to a preset improved AlexNet network model to determine the defect information on the vehicle brake disc; where the defect information includes the defect type and the quantity corresponding to the defect type; A marking module, configured to mark the vehicle brake disc according to the defect type and the quantity of the defect type to judge the quality of the vehicle brake disc; A judgment module, configured to confirm whether the production line needs to be shut down for maintenance according to the serial number of the vehicle brake disc, the defect information and the mark of the vehicle brake disc.

[0016] The automotive brake disc defect detection system according to an embodiment of the present invention enhances the surface image by using an improved Retinex algorithm, effectively improving the overall contrast of the surface image, making the local features of the automotive brake disc more obvious, and providing a basic guarantee for improving the AlexNet network model for detection and recognition; by using the improved AlexNet network model, the learning ability of the model is improved, the model recognition speed is accelerated, and the defect detection of the surface image can achieve an efficient and accurate effect; through different defect types and the number of defect types, the automotive brake disc is analyzed and judged at multiple levels to mark the automotive brake discs that do not meet the production quality. Finally, according to the marking results and the defect information of the automotive brake disc, it is judged whether there are defects in the production line. If there are defects, the production line will be shut down for maintenance to avoid greater losses and effectively reduce the production cost.

[0017] Further, the obtaining module is specifically configured to: Replace the logarithmic transformation in the Retinex algorithm with an adaptive power transformation to obtain an initial enhanced image:

[0018] Wherein, is the initial enhanced image, is the original image, is the Gaussian low-pass filter, is the adaptive power exponent, , is the information entropy of the surface image, are the row and column of the image matrix respectively; Use a sliding window of size to traverse the initial enhanced image , obtain the sliding window average value and the sliding window variance, and obtain the target enhanced image according to the sliding window average value and the sliding window variance, thereby completing the image enhancement preprocessing of the surface image; Wherein, the sliding window average value, the sliding window variance and the target enhanced image are:

[0019]

[0020]

[0021] Wherein, is the sliding window average value, is the sliding window variance, is the target enhanced image.

[0022] Further, the improved AlexNet network model includes 6 convolutional layers, 4 pooling layers and 3 fully connected layers, and the BN layer is used to replace the LRN normalization layer in the traditional AlexNet network model; Among them, the convolution kernel size of the Conv1 convolutional layer is 13×13, the convolution kernel size of the Conv2 convolutional layer is 5×5, the convolution kernel sizes of the Conv3 convolutional layer, the Conv4 convolutional layer and the Conv5 convolutional layer are all 3×3, and the convolution kernel size of the Conv6 convolutional layer is 1×1.

[0023] Further, the marking module is specifically used for: Classify the defects of the automotive brake disc into hole defects and line defects, and record the automotive brake disc after being detected and recognized by the improved AlexNet network model as The The number of the hole defects is recorded as The The number of the line defects is recorded as where is the number of the automotive brake disc; If When, then mark the automotive brake disc with the number as a defective product; where is the defect quantity threshold; Calibrate the defects of the automotive brake disc based on the pixel equivalent of shape matching to obtain the sizes of the defects of the automotive brake disc, which are respectively recorded as and where is the size of the th hole defect of the automotive brake disc with the number is the size of the th line defect of the automotive brake disc with the number If or any one of the conditions is satisfied, then mark the automotive brake disc with the number as a defective product; where is the first size threshold of the hole defect, is the first size threshold of the line defect; If or any one of the conditions is satisfied, then mark the automotive brake disc with the number as a defective product; where is the second size threshold of the hole defect, is the second size threshold of the line defect.

[0024] Further, the determination module is specifically configured to: Record the quality mark of the vehicle brake disc numbered as , where when, the vehicle brake disc numbered is a defective product, when, the vehicle brake disc numbered is a qualified product; If , it is determined that the production line needs to be shut down for maintenance; If , and , and at the same time at least one of them is 0, it is determined that the production line needs to be shut down for maintenance; where is the number of hole defects of the vehicle brake disc numbered , is the number of hole defects of the vehicle brake disc numbered , is the number of line defects of the vehicle brake disc numbered , is the number of line defects of the vehicle brake disc numbered . BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0026] Figure 1 is a flowchart of the vehicle brake disc defect detection method according to Embodiment 1 of the present invention; Figure 2 is a schematic structural diagram of the improved AlexNet network model in the vehicle brake disc defect detection method according to Embodiment 1 of the present invention; Figure 3 is a structural block diagram of the vehicle brake disc defect detection system according to Embodiment 2 of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] ​​​Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the embodiments of the present invention and should not be construed as limiting the present invention.

[0028] In the description of the embodiments of the present invention, it should be understood that the orientation or positional relationships indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. are based on the orientation or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as limiting the present invention.

[0029] In addition, the terms "first" and "second" are only used for descriptive purposes and should not be construed as indicating or implying relative importance or implicitly indicating the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include one or more of such features. In the description of the embodiments of the present invention, the meaning of "a plurality" is two or more unless otherwise specifically defined.

[0030] In the embodiments of the present invention, unless otherwise clearly specified and limited, the terms "mounted", "connected", "connected to", "fixed", etc. should be understood in a broad sense. For example, it may be a fixed connection, a detachable connection, or integrated; it may be a mechanical connection or an electrical connection; it may be directly connected or indirectly connected through an intermediate medium, and it may be the internal communication of two elements or the interaction relationship between two elements. For those of ordinary skill in the art, the specific meanings of the above terms in the embodiments of the present invention can be understood according to specific circumstances.

[0031] Embodiment 1 Referring to Figures 1 to 2 , in the first embodiment of the present invention, a method for detecting defects in an automotive brake disc, the method includes: S1: Obtain the number and surface image of the automotive brake disc, and perform image enhancement preprocessing on the surface image based on an improved Retinex algorithm. Further, the step of performing image enhancement preprocessing on the surface image based on the improved Retinex algorithm specifically includes: Replace the logarithmic transformation in the Retinex algorithm with an adaptive power transformation to obtain an initial enhanced image:

[0032] Wherein, is the initial enhanced image, is the original image, is the Gaussian low-pass filter, is the adaptive power exponent, , is the information entropy of the surface image, are the row and column of the image matrix respectively; effectively enhancing the overall contrast of the image; Adopt a sliding window of size to traverse the initial enhanced image , obtaining the sliding window average value and the sliding window variance, and obtaining the target enhanced image according to the sliding window average value and the sliding window variance, thereby completing the image enhancement preprocessing of the surface image; further improving the visual effect of the image and making the local features of the image more obvious; Among them, the sliding window average value, the sliding window variance and the target enhanced image are:

[0033]

[0034]

[0035] Among them, is the sliding window average value, is the sliding window variance, is the target enhanced image.

[0036] S2: Detect and identify the preprocessed surface image according to the preset improved AlexNet network model to determine the defect information on the automotive brake disc; among them, the defect information includes the defect type and the quantity corresponding to the defect type; further, the improved AlexNet network model includes 6 convolutional layers, 4 pooling layers and 3 fully connected layers, and uses the BN layer to replace the LRN normalization layer in the traditional AlexNet network model; Among them, the convolutional kernel size of the Conv1 convolutional layer is 13×13, the convolutional kernel size of the Conv2 convolutional layer is 5×5, the convolutional kernel sizes of the Conv3 convolutional layer, the Conv4 convolutional layer and the Conv5 convolutional layer are all 3×3, and the convolutional kernel size of the Conv6 convolutional layer is 1×1.

[0037] The improved AlexNet network model adds a convolutional layer compared to the traditional AlexNet network model, and expands the size of the convolutional kernel, improving the global feature recognition effect of the model. The BN layer can reduce the impact caused by parameter changes in the network, ensure the capacity of the network, accelerate the network convergence speed, further improve the performance of the model, and ensure the detection and recognition speed and accuracy of the model.

[0038] S3: Mark the automotive brake disc according to the defect type and the quantity of the defect type to judge the quality of the automotive brake disc; Further, the step of marking the automotive brake disc according to the defect type and the quantity of the defect type to judge the quality of the automotive brake disc specifically includes: Classify the defects of the automotive brake disc into hole defects and line defects, and denote the automotive brake disc after being detected and recognized by the improved AlexNet network model as , the quantity of the hole defects is denoted as , the quantity of the line defects is denoted as , where is the number of the automotive brake disc; If when, then mark the automotive brake disc with the number as a defective product; where is the defect quantity threshold; In this embodiment, ; Calibrate the defects of the automotive brake disc based on the pixel equivalent of shape matching to obtain the sizes of the defects of the automotive brake disc, which are respectively denoted as and , where is the size of the th hole defect of the automotive brake disc with the number , is the size of the th line defect of the automotive brake disc with the number ; If or any one of the conditions is satisfied, then mark the automotive brake disc with the number as a defective product; where is the first size threshold of the hole defect, is the first size threshold of the line defect; In this implementation, is the radius of the hole defect, , is the length of the line defect, ; If the following conditions are met or any one of the conditions is satisfied, then the automobile brake disc numbered is a defective product; where is the second size threshold of the hole defect is the second size threshold of the line defect; in this embodiment , .

[0039] S4: Determine whether the production line needs to stop production for maintenance according to the number of the automobile brake disc, the defect information and the mark of the automobile brake disc; further, the steps of determining whether the production line needs to stop production for maintenance according to the number of the automobile brake disc, the defect information and the mark of the automobile brake disc specifically include: Record the quality mark of the automobile brake disc numbered as , where when the automobile brake disc numbered is a defective product when the automobile brake disc numbered is a qualified product; If , then it is determined that the production line needs to stop production for maintenance; that is, three consecutive automobile brake discs are defective products. At this time, it is judged that there is a problem with the production line and it is necessary to stop production for maintenance immediately to avoid greater losses; If , and , and at the same time at least one of them is 0, then it is determined that the production line needs to stop production for maintenance; where is the number of hole defects of the automobile brake disc numbered is the number of hole defects of the automobile brake disc numbered is the number of line defects of the automobile brake disc numbered is the number of line defects of the automobile brake disc numbered ; at this time, the product quality is in a downward state, and it is inferred that there are defects in the production line process, and it is necessary to adjust each process parameter to avoid continuous decline of product quality.

[0040] The method for detecting defects of an automobile brake disc according to an embodiment of the present invention enhances the surface image through the improved Retinex algorithm, effectively improving the overall contrast of the surface image, making the local features of the automobile brake disc more obvious, and providing a basic guarantee for improving the AlexNet network model for detection and recognition; by using the improved AlexNet network model, the learning ability of the model is improved, the model recognition speed is accelerated, and the defect detection of the surface image can achieve efficient and accurate results; through different defect types and the number of defect types, the automobile brake disc is analyzed and judged at multiple levels to mark the automobile brake disc that does not meet the production quality. Finally, according to the marking result and the defect information of the automobile brake disc, it is judged whether there are defects in the production line. If so, production is stopped for maintenance to avoid greater losses and effectively reduce the production cost.

[0041] Embodiment 2 Refer to Figure 3 , the present invention also provides an automobile brake disc defect detection system, which includes: An acquisition module, configured to acquire the number and surface image of the automobile brake disc, and perform image enhancement preprocessing on the surface image based on the improved Retinex algorithm; A detection module, configured to perform detection and recognition on the preprocessed surface image according to a preset improved AlexNet network model to determine the defect information on the automobile brake disc; wherein, the defect information includes the defect type and the quantity corresponding to the defect type; A marking module, configured to mark the automobile brake disc according to the defect type and the quantity of the defect type to judge the quality of the automobile brake disc; A judgment module, configured to confirm whether the production line needs to stop production and be overhauled according to the number of the automobile brake disc, the defect information and the marking of the automobile brake disc.

[0042] The automobile brake disc defect detection system according to an embodiment of the present invention enhances the surface image through the improved Retinex algorithm, effectively improving the overall contrast of the surface image, making the local features of the automobile brake disc more obvious, and providing a basic guarantee for improving the AlexNet network model for detection and recognition; by using the improved AlexNet network model, the learning ability of the model is improved, the model recognition speed is accelerated, and the defect detection of the surface image can achieve efficient and accurate results; through different defect types and the number of defect types, the automobile brake disc is analyzed and judged at multiple levels to mark the automobile brake disc that does not meet the production quality. Finally, according to the marking result and the defect information of the automobile brake disc, it is judged whether there are defects in the production line. If so, production is stopped for maintenance to avoid greater losses and effectively reduce the production cost.

[0043] Further, the obtaining module is specifically configured to: Replace the logarithmic transformation in the Retinex algorithm with an adaptive power transformation to obtain an initial enhanced image:

[0044] where, is the initial enhanced image, is the original image, is the Gaussian low-pass filter, is the adaptive power exponent, , is the information entropy of the surface image, are the row and column of the image matrix respectively; Use a sliding window of size to traverse the initial enhanced image , obtain the sliding window average value and the sliding window variance, and obtain the target enhanced image according to the sliding window average value and the sliding window variance, so as to complete the image enhancement preprocessing of the surface image; where, the sliding window average value, the sliding window variance and the target enhanced image are:

[0045]

[0046]

[0047] where, is the sliding window average value, is the sliding window variance, is the target enhanced image.

[0048] Further, the improved AlexNet network model includes 6 convolutional layers, 4 pooling layers and 3 fully connected layers, and uses the BN layer to replace the LRN normalization layer in the traditional AlexNet network model; Among them, the convolution kernel size of the Conv1 convolutional layer is 13×13, the convolution kernel size of the Conv2 convolutional layer is 5×5, the convolution kernel sizes of the Conv3 convolutional layer, the Conv4 convolutional layer and the Conv5 convolutional layer are all 3×3, and the convolution kernel size of the Conv6 convolutional layer is 1×1.

[0049] Further, the marking module is specifically configured to: Classify the defects of the automotive brake disc into hole defects and line defects, and record the automotive brake disc after being detected and recognized by the improved AlexNet network model as , the The number of the hole defects is denoted as , the The number of the line defects is denoted as , where is the serial number of the automotive brake disc; If , then the automotive brake disc with the serial number is marked as a defective product; where is the defect quantity threshold; Based on the pixel equivalent of shape matching, the defects of the automotive brake disc are calibrated to obtain the sizes of the defects of the automotive brake disc, which are respectively denoted as and , where is the size of the th hole defect of the automotive brake disc with the serial number , is the size of the th line defect of the automotive brake disc with the serial number ; If or is satisfied, then the automotive brake disc with the serial number is marked as a defective product; where is the first size threshold of the hole defect, is the first size threshold of the line defect; If or is satisfied, then the automotive brake disc with the serial number is marked as a defective product; where is the second size threshold of the hole defect, is the second size threshold of the line defect.

[0050] Furthermore, the judgment module is specifically configured to: Denote the quality mark of the automotive brake disc with the serial number as , where , the automotive brake disc with the serial number is a defective product, , the automotive brake disc with the serial number is a qualified product; If , then it is determined that the production line needs to stop production for maintenance; If , and , and at the same time at least one of them is 0, then it is determined that the production line needs to stop production for maintenance; where is the number of hole defects of the automotive brake disc numbered , and is the number of hole defects of the automotive brake disc numbered . is the number of line defects of the automotive brake disc numbered , and is the number of line defects of the automotive brake disc numbered . Embodiment III In the third embodiment of the present invention, based on the same inventive concept, a computer-readable storage medium is proposed. The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the automotive brake disc defect detection method in the above embodiment are implemented. The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a definite sequence list of executable instructions for implementing a logical function, and can be specifically implemented in any computer-readable medium for use by an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processor, or other systems that can fetch instructions from the instruction execution system, apparatus, or device and execute the instructions), or in combination with these instruction execution systems, apparatus, or devices. For the purposes of this specification, a "computer-readable medium" can be any device that includes, stores, communicates, propagates, or transports a program for use by or in connection with an instruction execution system, apparatus, or device. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection part (electronic device) having one or more wirings, a portable computer disk cartridge (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium can even be paper or other suitable media on which the program can be printed, because the program can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpretation, or other suitable processing as necessary, and then stored in a computer memory.

[0051] Among them, the memory may include a mass storage for data or instructions. By way of example and not limitation, the memory may include a hard disk drive (HDD), a floppy disk drive, a solid state drive (SSD), a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive, or a combination of two or more of these. Where appropriate, the memory may include removable or non-removable (or fixed) media. Where appropriate, the memory may be internal or external to the data processing device. In a particular embodiment, the memory is non-volatile memory. In a particular embodiment, the memory includes a read-only memory (ROM) and a random access memory (RAM). Where appropriate, the ROM may be a mask-programmed ROM, a programmable ROM (PROM), an erasable PROM (EPROM), an electrically erasable PROM (EEPROM), an electrically alterable ROM (EAROM), or a flash memory, or a combination of two or more of these. Where appropriate, the RAM may be a static random access memory (SRAM) or a dynamic random access memory (DRAM), where the DRAM may be a fast page mode dynamic random access memory (FPMDRAM), an extended date out dynamic random access memory (EDODRAM), a synchronous dynamic random access memory (SDRAM), etc.

[0052] Embodiment Four The fourth embodiment of the present invention. Based on the same inventive concept, a terminal proposed by the present invention includes: a processor and a memory; the processor and the memory communicate with each other; the memory is used to store instructions; the processor is used to execute the instructions in the memory to execute the automotive brake disc defect detection method of the above embodiment.

[0053] It should be understood that each part of the present invention can be implemented by hardware, software, firmware, or a combination thereof. In the above embodiment, multiple steps or methods can be implemented by software or firmware stored in the memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following well-known technologies in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0054] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0055] On the premise of no conflict, those skilled in the art can freely combine and superimpose the above additional technical features.

[0056] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting defects in automobile brake discs, characterized in that: The method comprises: Obtaining the serial number and surface image of the automobile brake disc, and performing image enhancement preprocessing on the surface image based on the improved Retinex algorithm; Detecting and identifying the preprocessed surface image according to a preset improved AlexNet network model to determine defect information on the automobile brake disc; wherein the defect information includes a defect type and a quantity corresponding to the defect type; Marking the automobile brake disc according to the defect type and the number of the defect type to determine the quality of the automobile brake disc; Determine whether the production line needs to be shut down for maintenance based on the number of the automobile brake disc, the defect information and the mark of the automobile brake disc.

2. The automobile brake disc defect detection method according to claim 1, characterized in that: The step of performing image enhancement preprocessing on the surface image based on the improved Retinex algorithm specifically includes: The adaptive power transform replaces the logarithmic transform in the Retinex algorithm to obtain the initial enhanced image: in, is the initial enhanced image, is the original image, is a Gaussian low-pass filter, is the adaptive power exponent, , is the information entropy of the surface image, are the rows and columns of the image matrix respectively; The size is The sliding window traverses the initial enhanced image , obtaining a sliding window average and a sliding window variance, and obtaining a target enhanced image according to the sliding window average and the sliding window variance, thereby completing image enhancement preprocessing of the surface image; Among them, the sliding window mean, the sliding window variance and the target enhanced image are: in, is the sliding window average value, is the sliding window variance, An image is enhanced for the target.

3. The automobile brake disc defect detection method according to claim 1, characterized in that: The improved AlexNet network model includes 6 convolutional layers, 4 pooling layers and 3 fully connected layers, and a BN layer is used to replace the LRN normalization layer in the traditional AlexNet network model; Among them, the convolution kernel size of the Conv1 convolution layer is 13×13, the convolution kernel size of the Conv2 convolution layer is 5×5, the convolution kernel sizes of the Conv3 convolution layer, Conv4 convolution layer and Conv5 convolution layer are all 3×3, and the convolution kernel size of the Conv6 convolution layer is 1×1.

4. The automobile brake disc defect detection method according to claim 1, characterized in that: The step of marking the automobile brake disc according to the defect type and the number of the defect type to judge the quality of the automobile brake disc specifically includes: The defects of the automobile brake disc are divided into hole defects and line defects, and the automobile brake disc detected and identified by the improved AlexNet network model is recorded as , The number of the hole defects is recorded as , The number of line defects is recorded as ,in, The serial number of the automobile brake disc; like , then the tag number is The automobile brake disc is a defective product; wherein, is the defect quantity threshold; The defect of the automobile brake disc is calibrated based on the pixel equivalent of shape matching, and the size of the defect of the automobile brake disc is obtained, which is recorded as and ,in, For the number The automobile brake disc The size of the hole defect, For the number The automobile brake disc The size of the line defect; If satisfied or Any condition, then the tag number is The automobile brake disc is a defective product; wherein, is the first size threshold of the hole defect, is a first size threshold of the line defect; If satisfied or Any condition, then the tag number is The automobile brake disc is a defective product; wherein, is the second size threshold of the hole defect, is a second size threshold of the line defect.

5. The automobile brake disc defect detection method according to claim 4, characterized in that: The steps of confirming whether the production line needs to be stopped for maintenance according to the number of the automobile brake disc, the defect information and the mark of the automobile brake disc specifically include: Record number as The quality mark of the automobile brake disc is ,in, When The automobile brake disc is a defective product. When The automobile brake disc is a qualified product; like , it is determined that the production line needs to be shut down for maintenance; like ,and ,at the same time If at least one of them is 0, it is determined that the production line needs to be shut down for maintenance; For the number The number of hole defects of the automobile brake disc, For the number The number of hole defects of the automobile brake disc, For the number The number of line defects of the automobile brake disc, For the number The number of line defects on the automobile brake disc.

6. A vehicle brake disc defect detection system, characterized in that: The system comprises: An acquisition module, used for acquiring the serial number and surface image of the automobile brake disc, and performing image enhancement preprocessing on the surface image based on an improved Retinex algorithm; A detection module, used to detect and identify the preprocessed surface image according to a preset improved AlexNet network model, and determine the defect information on the automobile brake disc; wherein the defect information includes the defect type and the quantity corresponding to the defect type; A marking module, used for marking the automobile brake disc according to the defect type and the quantity of the defect type to judge the quality of the automobile brake disc; The judgment module is used to confirm whether the production line needs to be stopped for maintenance based on the number of the automobile brake disc, the defect information and the mark of the automobile brake disc.

7. The automobile brake disc defect detection system according to claim 6, characterized in that: The acquisition module is specifically used for: The adaptive power transform replaces the logarithmic transform in the Retinex algorithm to obtain the initial enhanced image: in, is the initial enhanced image, is the original image, is a Gaussian low-pass filter, is the adaptive power exponent, , is the information entropy of the surface image, are the rows and columns of the image matrix respectively; The size is The sliding window traverses the initial enhanced image , obtaining a sliding window average and a sliding window variance, and obtaining a target enhanced image according to the sliding window average and the sliding window variance, thereby completing image enhancement preprocessing of the surface image; Among them, the sliding window mean, the sliding window variance and the target enhanced image are: in, is the sliding window average value, is the sliding window variance, An image is enhanced for the target.

8. The automobile brake disc defect detection system according to claim 6, characterized in that: The improved AlexNet network model includes 6 convolutional layers, 4 pooling layers and 3 fully connected layers, and a BN layer is used to replace the LRN normalization layer in the traditional AlexNet network model; Among them, the convolution kernel size of the Conv1 convolution layer is 13×13, the convolution kernel size of the Conv2 convolution layer is 5×5, the convolution kernel sizes of the Conv3 convolution layer, Conv4 convolution layer and Conv5 convolution layer are all 3×3, and the convolution kernel size of the Conv6 convolution layer is 1×1.

9. The automobile brake disc defect detection system according to claim 6, characterized in that: The marking module is specifically used for: The defects of the automobile brake disc are divided into hole defects and line defects, and the automobile brake disc detected and identified by the improved AlexNet network model is recorded as , The number of the hole defects is recorded as , The number of line defects is recorded as ,in, The serial number of the automobile brake disc; like , then the tag number is The automobile brake disc is a defective product; wherein, is the defect quantity threshold; The defect of the automobile brake disc is calibrated based on the pixel equivalent of shape matching, and the size of the defect of the automobile brake disc is obtained, which is recorded as and ,in, For the number The automobile brake disc The size of the hole defect, For the number The automobile brake disc The size of the line defect; If satisfied or Any condition, then the tag number is The automobile brake disc is a defective product; wherein, is the first size threshold of the hole defect, is a first size threshold of the line defect; If satisfied or Any condition, then the tag number is The automobile brake disc is a defective product; wherein, is the second size threshold of the hole defect, is a second size threshold of the line defect.

10. The automobile brake disc defect detection system according to claim 9, characterized in that: The judgment module is specifically used for: Record number as The quality mark of the automobile brake disc is ,in, When The automobile brake disc is a defective product. When The automobile brake disc is a qualified product; like , it is determined that the production line needs to be shut down for maintenance; like ,and ,at the same time If at least one of them is 0, it is determined that the production line needs to be shut down for maintenance; For the number The number of hole defects of the automobile brake disc, For the number The number of hole defects of the automobile brake disc, For the number The number of line defects of the automobile brake disc, For the number The number of line defects on the automobile brake disc.

Images