PLC control method based on polarization image denoising and demosaicing

Abstract

The invention discloses a PLC control method based on polarization image denoising and demosaicing, and relates to the technical field of PLC control, the PLC control method comprises a PLC configuration stage, image acquisition and processing and PLC remote control, after detailed configuration and debugging are carried out on a PLC and all related equipment, corresponding processing is carried out on an acquired polarization image, and the polarization image denoising and demosaicing can be realized. The method comprises the following steps: processing polarization images to obtain processed polarization images, processing the polarization images based on loss functions, balancing different types of losses, and executing operation control of equipment corresponding to a PLC, so that the problem of fusion of a polarization image processing technology and a PLC control system is solved to a certain extent, and the problems of polarization image noise interference and mosaic effect are solved; according to the invention, visual remote control of industrial equipment is realized, noise interference in a polarization image can be effectively eliminated, details of the polarization image are clearer and finer, the color rendition degree and the spatial resolution of the polarization image are improved, and the accuracy of image colors is ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of PLC control, and in particular to a PLC control method based on polarization image denoising and demosaicing. Background Art

[0002] With the continuous advancement of industrial automation and intelligent manufacturing, image processing technology is increasingly used in industrial monitoring and quality control. Polarization cameras can obtain rich surface features by capturing reflected light information on the surface of objects, and can provide high-precision and reliable multi-faceted functional experience in fields such as object recognition and surface defect detection. Based on the integration of polarization image processing technology and PLC control systems, image display in actual control scenarios can be quickly synchronized with control instructions, ensuring the accuracy of image color.

[0003] Prior art, such as a polarization image denoising method and its application disclosed in the invention patent application with announcement number CN118469860A, firstly, a noisy polarization image is decomposed into low-frequency components and high-frequency components using a multi-scale Gaussian filter; secondly, LF-UNet and HF-UNet are used to extract the features of multi-scale low-frequency and high-frequency components respectively; thirdly, a multi-scale two-level fusion strategy is established; finally, residual prediction is performed using high-frequency and low-frequency enhancement features, and then it is fused with the input image to obtain a high-quality polarization image. The present invention effectively decomposes and multi-scales the polarization image under haze weather, and uses the denoised polarization image for orientation, thereby improving the ability of atmospheric polarization light to resist weather interference during orientation, and realizing the accurate orientation of polarization light navigation under haze weather, which can be used for vehicle navigation and positioning in haze environment and other occasions.

[0004] In view of the above scheme, the applicant of the present invention has found that the above technical problems have at least the following technical problems: the above invention mainly decomposes and multi-scales the polarization image under haze weather, and does not combine the polarization image with PLC control to perform corresponding polarization image denoising and demosaicing, and does not perform corresponding configuration and debugging of PLC and related equipment. It is impossible to understand the current status of PLC and related equipment, and the accuracy of control cannot be guaranteed. It is impossible to ensure that PLC can correctly receive and process the corresponding output signal. At the same time, the polarization image collected by the polarization camera is not processed in each step, and the consistency of the final result image presentation cannot be guaranteed. It will greatly reduce the accuracy of the image when performing denoising and demosaicing, making the details of the polarization image blurred, and cannot lay an effective information data benchmark for the final PLC operation control. It cannot improve the color reproduction and spatial resolution of the polarization image, and reduce the accuracy of the image color. Summary of the invention

[0005] In view of the above-mentioned technical deficiencies, an object of the present invention is to provide a PLC control method based on polarization image denoising and demosaicing.

[0006] To solve the above technical problems, the present invention adopts the following technical solution: The present invention provides a PLC control method based on polarization image denoising and demosaicing, including: Step 1, PLC configuration stage: through detailed configuration and debugging of PLC and related equipment.

[0007] Step 2: Image acquisition and processing: The corresponding polarization image is acquired based on the polarization camera, and the acquired polarization image is processed accordingly to extract the corresponding polarization features in the polarization image to generate a polarization spectral attention image. At the same time, based on the extracted spectral features, a polarization result image is generated and weighted combination is performed based on the corresponding function method.

[0008] Step 3: PLC remote control: According to the weighted polarization information, the operation control of the corresponding PLC equipment is executed.

[0009] Compared with the prior art, the beneficial effects of the present invention are: 1. The present invention provides a PLC control method based on polarization image denoising and de-mosaicing. After detailed configuration and debugging of the PLC and related equipment, corresponding processing is performed on the collected polarization images to obtain each processed polarization image, and the polarization image is processed based on each loss function to balance different types of losses, so as to perform operation control of the corresponding PLC equipment. To a certain extent, the integration problem of polarization image processing technology and PLC control system is solved, and the problems of polarization image noise interference and mosaic effect are solved, and visual remote control of industrial equipment is realized. It can effectively eliminate noise interference in polarization images, make the details of polarization images clearer and more delicate, solve the shortcomings of current technology, improve the color reproduction and spatial resolution of polarization images, ensure the accuracy of image colors, and also reduce the misoperation caused by image analysis errors to a certain extent, improve safety and reliability, improve the efficiency of rapid recognition of image features, and realize more efficient image problem recognition and analysis.

[0010] 2. Based on detailed configuration and debugging of PLC and related equipment, efficient association between PLC and polarization image is achieved, effective control of PLC is guaranteed, stability and response speed of PLC under different working conditions are ensured, and effective fusion of PLC and polarization image is achieved.

[0011] 3. According to the corresponding polarization image collected, the corresponding various processing operations are performed to obtain the final standardized polarization image data, and the corresponding loss weighting is performed on it to ensure the loss balance of the image. To a certain extent, the noise interference in the polarization image is effectively eliminated, and the details of the polarization image are clearer and more delicate. The color reproduction and spatial resolution of the polarization image are improved, and the problems of noise interference and mosaic effect of the polarization image are efficiently solved. The effective control of PLC is realized, and the color accuracy of the polarization image is guaranteed. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. 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 paying creative work.

[0013] Figure 1 The present invention is a schematic flow chart of the steps for implementing the method. DETAILED DESCRIPTION

[0014] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0015] See also Figure 1 As shown, a PLC control method based on polarization image denoising and demosaicing includes: step one, PLC configuration stage: through detailed configuration and debugging of PLC and related equipment.

[0016] It should be noted that physical quantities include temperature, pressure, humidity, etc.

[0017] It should be noted that the sensors include temperature sensors, pressure sensors, humidity sensors, etc.

[0018] It should be noted that the relevant parameters of the measuring points are input into the PLC system to ensure that the PLC system can correctly receive and process these signals; a unique IP address is assigned to ensure that they can communicate with each other through the network; the communication parameters of the PLC are set to ensure smooth connection with remote devices, sensors and host computers; the communication protocol configuration is to ensure the compatibility between the PLC module and polarization cameras, sensors and other equipment, and to ensure that the PLC can exchange data with the host computer, server and other devices through the protocol.

[0019] As an optional implementation, first determine the location of the measuring points of the physical quantities that need to be monitored on site, then set corresponding sensors for each measuring point, configure the sensor type, measurement range and accuracy requirements, and input the relevant parameters of each measuring point monitored by the sensor into the PLC system.

[0020] Secondly, determine the network topology of the PLC, host computer, server and polarization camera corresponding to each device, assign unique IP addresses to the PLC and each device, and set the communication parameters of the PLC, including the subnet mask and gateway address.

[0021] Finally, select the corresponding communication protocol according to the system requirements, including Modbus, TCP / IP and Profinet, and configure the communication protocol parameters on the PLC, including baud rate, data bit, check bit and stop bit. Based on the above, complete the detailed configuration of the PLC and related equipment.

[0022] As an optional implementation, the PLC and related equipment are debugged, and the specific debugging process is as follows: according to the specific application requirements, the PLC control logic is written, including the collection, processing and feedback control of the data of each measuring point, the collection, data analysis and feedback control logic of the sensor signal are implemented in the PLC, and the control instructions are triggered according to the collected data, so as to debug the PLC program to see whether the sensors on site respond correctly. If the sensor shows a correct response state, the other related on-site equipment is controlled. Otherwise, if not, the corresponding debugging adjustments are performed to debug the PLC and related equipment accordingly.

[0023] It should be noted that the other on-site devices include motors and switches.

[0024] Based on detailed configuration and debugging of PLC and related equipment, efficient association between PLC and polarization image is achieved, effective control of PLC is guaranteed, stability and response speed of PLC under different working conditions are ensured, and effective fusion of PLC and polarization image is realized.

[0025] Step 2: Image acquisition and processing: The corresponding polarization image is acquired based on the polarization camera, and the acquired polarization image is processed accordingly to extract the corresponding polarization features in the polarization image to generate a polarization spectral attention image. At the same time, based on the extracted spectral features, a polarization result image is generated and weighted combination is performed based on the corresponding function method.

[0026] It should be noted that the image information includes polarization angle, polarization degree, light intensity, RGB value, roughness and direction, etc.

[0027] As an optional implementation, the corresponding polarization image is collected based on the polarization camera, and the specific collection process is as follows: B1, by remotely monitoring the location of the scene or object according to the corresponding polarization camera, and then using the polarization camera to capture the corresponding image information of the scene in real time, and then based on the PLC, the captured image information data is packaged and uploaded to the network server; B2. At the same time, each remote sensor monitors the physical quantity of each measuring point, and the sensor collects the measurement value of each monitoring point in real time to generate an analog signal. The sensor is connected to the PLC. When the analog signal is converted into digital data, it is uploaded to the server via the network. Based on the above, the corresponding polarization image is collected.

[0028] As an optional implementation, the corresponding processing is performed on the collected polarization image, and the specific processing process is as follows: S1. Perform corresponding preprocessing on the selected polarization image, first decompose the polarization image into three independent color channels, namely red, green and blue, and optimize the processing according to the characteristics of each color channel.

[0029] S2, decompose the polarization image into , and Three independent channel images to obtain multi-channel polarization images ,in include , , .

[0030] S3, multi-channel polarization image The independent channel images are denoised by first normalizing each independent channel image and selecting the corresponding denoising algorithm to denoise each independent channel image, maintaining the spatial resolution of the polarization image, capturing the features of the image at different scales, and dynamically adjusting each channel or spatial position. When the denoising is completed, the denoised multi-channel polarization image is Combine the denoised polarization images by stitching the independent color channels , thereby completing the processing corresponding to the polarization image.

[0031] It should be noted that the denoised multi-channel polarization image Combine the denoised polarization images by stitching the independent color channels : Multi-channel polarization image is stitched together through color channel operation ,Right now , , Merge into denoised polarization image .

[0032] It should also be noted that the optimization processing includes but is not limited to histogram equalization, gamma correction, filtering, edge sharpening and other technologies, and the visual effects of each channel are enhanced based on the optimization processing technology; the characteristics of each color channel include brightness and contrast, color saturation and noise level, etc.

[0033] It should also be noted that denoising algorithms include spatial domain filtering, frequency domain filtering, wavelet transform and deep learning denoising.

[0034] As an optional implementation, the extraction of the corresponding polarization features in the polarization image is specifically performed as follows: the denoised polarization image Input to the remote server to generate the corresponding polarization feature image , and generate polarization feature images The corresponding polarization feature extraction is performed, which is divided into three parts, including convolution calculation, activation function and normalization. First, the distribution, change and spatial structure of polarization features of each polarization information data in each color channel in the polarization image are captured. Secondly, based on the enhanced network's ability to express demosaic features, the feature changes corresponding to each polarization information data are captured. Finally, the normalization method is used to reduce the sensitivity of the demosaic model to the selection of initialization and learning rate, so as to extract the corresponding polarization features in the polarization image, including polarization degree and polarization angle. The polarization degree calculation formula is: , is the parallel of polarized light intensity, is the vertical component of polarized light intensity, and the polarization angle calculation formula is: .

[0035] As an optional implementation, the polarization spectrum attention image is generated, and the specific generation process is as follows: the polarization feature image Polarized spectral attention image generated as input , and utilize the complementarity of polarization spatial information and spectral information and the attention mechanism to enhance important features and suppress irrelevant information. By introducing various convolutional layers and activation functions, the spatial and spectral features of the polarization image are preliminarily extracted, where the spectral features include scattering characteristics, optical characteristics, and reflection characteristics. The spatial attention map is generated according to the additional convolutional layers. At the same time, based on the importance of a certain area in the polarization image for the demosaicing task, the spatial attention map is multiplied with the original feature map to enhance important areas and suppress non-important areas. Finally, the spectral attention is applied to each channel feature in a weighted summation manner to obtain the spectral feature information after the fusion of polarization information and spectral information, and a polarization spectral attention image is generated.

[0036] As an optional implementation, the polarization result image is generated, and the specific generation process is as follows: the polarization spectrum attention image Generates a polarization result image as input Based on the spectral feature information after fusion in the polarization spectrum attention image, the corresponding mosaic optimization is performed, and the noise and artifacts in the polarization spectrum attention image are removed by using the nonlinear filtering algorithm. At the same time, the input polarization spectrum attention image is processed according to the convolution kernel operation. Local feature extraction, including edges and textures, is performed, and robust feature representation is performed during the deep learning process to comprehensively generate the polarization result image.

[0037] It should be noted again that the convolution kernel includes edge detection convolution kernel and texture analysis convolution kernel, etc.

[0038] As an optional implementation, the weighted combination is performed based on the corresponding function method, and the specific combination process is as follows: first, a pixel-level mean square error loss function is set to reduce the difference between the result image and the true polarization image, so that the result image is visually close to the real image; secondly, a structural similarity loss function is set to measure the similarity of the image in brightness, contrast and structural information, and maintain the texture and edge information of the image when removing noise, and restore the high-frequency details and structural information in the image; finally, a polarization information preservation loss function is set to compare the distribution differences in polarization angle and polarization degree parameters between the real image and the result image, so that the model can retain and restore the detail data corresponding to the polarization information when removing noise, thereby performing weighted combination based on the corresponding comprehensive loss function.

[0039] According to the acquisition of the corresponding polarization image, the corresponding various processing operations are performed to obtain the final standardized polarization image data, and the corresponding loss weighting is performed on it to ensure the loss balance of the image. To a certain extent, the noise interference in the polarization image is effectively eliminated, and the details of the polarization image are clearer and delicate. The color reproduction and spatial resolution of the polarization image are improved, and the problems of noise interference and mosaic effect of the polarization image are efficiently solved. The effective control of PLC is realized, and the color accuracy of the polarization image is guaranteed.

[0040] Step 3: PLC remote control: According to the weighted polarization information, the operation control of the corresponding PLC equipment is executed.

[0041] It should be noted that the on-site equipment includes camera pans, switches, motors, and electric pumps.

[0042] As an optional implementation, the weighted polarization information is input to the server, and then the corresponding control is generated: F1, automatic control, according to the calculation results automatically generated by the server based on the polarization information data, the control command is automatically issued to the PLC to control the operating status of the field equipment; F2. Human control: by observing the image details presented in the polarization image, according to engineering experience and based on the PLC programming configuration software and the host computer, corresponding instructions are sent to the PLC to control the operating status of the field equipment. Based on the above, the operation control of the corresponding equipment of the PLC is executed.

[0043] The embodiment of the present invention performs detailed configuration and debugging on the PLC and related equipment, and then performs corresponding processing on the collected polarization images to obtain each processed polarization image, and processes the polarization image based on each loss function to balance different types of losses, so as to execute the operation control of the corresponding equipment of the PLC. To a certain extent, it solves the integration problem of polarization image processing technology and PLC control system, solves the problems of polarization image noise interference and mosaic effect, realizes visual remote control of industrial equipment, can effectively eliminate noise interference in polarization images, makes the details of polarization images clearer and more delicate, solves the shortcomings of current technology, improves the color reproduction and spatial resolution of polarization images, ensures the accuracy of image colors, and also reduces the misoperation caused by image analysis errors to a certain extent, improves safety and reliability, improves the efficiency of rapid recognition of image features, and realizes more efficient image problem recognition and analysis.

[0044] The above contents are merely examples and explanations of the concept of the present invention. Those skilled in the art may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they shall all fall within the protection scope of the present invention.

Claims

1. A PLC control method based on polarization image denoising and demosaicing, characterized in that: include: Step 1: PLC configuration stage: through detailed configuration and debugging of PLC and related equipment; Step 2, image acquisition and processing: the corresponding polarization image is acquired based on the polarization camera, and the acquired polarization image is processed accordingly, the corresponding polarization features in the polarization image are extracted, and the polarization spectrum attention image is generated. At the same time, based on the extracted spectral features, the polarization result image is generated, and weighted combination is performed based on the corresponding function method; Step 3: PLC remote control: According to the weighted polarization information, the operation control of the corresponding PLC equipment is executed.

2. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The PLC and related equipment are configured in detail. The specific configuration process is as follows: First, determine the location of the measuring points of the physical quantities that need to be monitored on site, then set the corresponding sensors for each measuring point, configure the sensor type, measurement range and accuracy requirements, and input the relevant parameters of each measuring point monitored by the sensor into the PLC system; Secondly, determine the network topology of the PLC, host computer, server and polarization camera corresponding to each device, assign unique IP addresses to the PLC and each device, and set the communication parameters of the PLC, including the subnet mask and gateway address; Finally, select the corresponding communication protocol according to the system requirements, including Modbus, TCP / IP and Profinet, and configure the communication protocol parameters on the PLC, including baud rate, data bit, check bit and stop bit. Based on the above, complete the detailed configuration of the PLC and related equipment.

3. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The PLC and related equipment are debugged, and the specific debugging process is as follows: According to the specific application requirements, write the PLC control logic, including the collection, processing and feedback control of the data of each measuring point, implement the collection, data analysis and feedback control logic of the sensor signal in the PLC, and trigger the control instructions according to the collected data to debug the PLC program to see whether the sensors on site respond correctly. If the sensor responds correctly, control the other related on-site equipment. Otherwise, perform the corresponding debugging and adjustment to debug the PLC and related equipment accordingly.

4. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The polarization image is collected based on the polarization camera, and the specific collection process is as follows: B1. By remotely monitoring the location of the scene or object according to the corresponding polarization camera, the polarization camera is used to capture the corresponding image information of the scene in real time, and then the captured image information data is packaged and uploaded to the network server based on PLC; B2. At the same time, each remote sensor monitors the physical quantity of each measuring point, and the sensor collects the measurement value of each monitoring point in real time to generate an analog signal. The sensor is connected to the PLC. When the analog signal is converted into digital data, it is uploaded to the server via the network. Based on the above, the corresponding polarization image is collected.

5. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The corresponding processing is performed on the collected polarization image, and the specific processing process is as follows: S1. Perform corresponding preprocessing on the selected polarization image, first decompose the polarization image into three independent color channels, namely red, green and blue, and perform optimization processing according to the characteristics of each color channel; S2, decompose the polarization image into , and Three independent channel images to obtain multi-channel polarization images ,in include , , ; S3, multi-channel polarization image The independent channel images are denoised by first normalizing each independent channel image and selecting the corresponding denoising algorithm to denoise each independent channel image, maintaining the spatial resolution of the polarization image, capturing the features of the image at different scales, and dynamically adjusting each channel or spatial position. When the denoising is completed, the denoised multi-channel polarization image is Combine the denoised polarization images by stitching the independent color channels , thereby completing the processing corresponding to the polarization image.

6. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The specific extraction process of extracting the corresponding polarization features in the polarization image is as follows: The denoised polarization image Input to the remote server to generate the corresponding polarization feature image , and generate polarization feature images The corresponding polarization feature extraction is performed, which is divided into three parts, including convolution calculation, activation function and normalization. First, the distribution, change and spatial structure of polarization features of each polarization information data in each color channel in the polarization image are captured. Secondly, based on the enhanced network's ability to express demosaic features, the feature changes corresponding to each polarization information data are captured. Finally, the normalization method is used to reduce the sensitivity of the demosaic model to the selection of initialization and learning rate, so as to extract the corresponding polarization features in the polarization image, including polarization degree and polarization angle. The polarization degree calculation formula is: , is the parallel of polarized light intensity, is the vertical component of polarized light intensity, and the polarization angle calculation formula is: .

7. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The polarization spectrum attention image is generated, and the specific generation process is as follows: Polarization feature image Polarized spectral attention image generated as input , and utilize the complementarity of polarization spatial information and spectral information and the attention mechanism to enhance important features and suppress irrelevant information. By introducing various convolutional layers and activation functions, the spatial and spectral features of the polarization image are preliminarily extracted, where the spectral features include scattering characteristics, optical characteristics, and reflection characteristics. The spatial attention map is generated according to the additional convolutional layers. At the same time, based on the importance of a certain area in the polarization image for the demosaicing task, the spatial attention map is multiplied with the original feature map to enhance important areas and suppress non-important areas. Finally, the spectral attention is applied to each channel feature in a weighted summation manner to obtain the spectral feature information after the fusion of polarization information and spectral information, and a polarization spectral attention image is generated.

8. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The polarization result image is generated, and the specific generation process is as follows: Polarization spectral attention image Generates a polarization result image as input Based on the spectral feature information after fusion in the polarization spectrum attention image, the corresponding mosaic optimization is performed, and the noise and artifacts in the polarization spectrum attention image are removed by using the nonlinear filtering algorithm. At the same time, the input polarization spectrum attention image is processed according to the convolution kernel operation. Local feature extraction, including edges and textures, is performed, and robust feature representation is performed during the deep learning process to comprehensively generate the polarization result image.

9. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The weighted combination is performed based on the corresponding function method, and the specific combination process is as follows: First, a pixel-level mean square error loss function is set to reduce the difference between the result image and the true polarization image, making the result image visually close to the real image. Secondly, a structural similarity loss function is set to measure the similarity of the images in brightness, contrast and structural information, and to maintain the texture and edge information of the image when removing noise, and to restore the high-frequency details and structural information in the image. Finally, a polarization information preservation loss function is set to compare the distribution differences in polarization angle and polarization degree parameters between the real image and the result image, so that the model can retain and restore the detailed data corresponding to the polarization information when removing noise, and thus perform weighted combination based on the corresponding comprehensive loss function.

10. A PLC control method based on polarization image denoising and demosaicing as claimed in claim 1, characterized in that: The operation control of the equipment corresponding to the execution PLC is as follows: The weighted polarization information is input to the server, and then the corresponding control is generated: F1, automatic control, according to the calculation results automatically generated by the server based on the polarization information data, the control command is automatically sent to the PLC to control the operating status of the field equipment; F2. Human control: by observing the image details presented in the polarization image, according to engineering experience and based on the PLC programming configuration software and the host computer, corresponding instructions are sent to the PLC to control the operating status of the field equipment. Based on the above, the operation control of the corresponding equipment of the PLC is executed.

Images