Computer image processing system based on intelligent identification technology

Abstract

The invention discloses a computer image processing system based on an intelligent identification technology, and the system comprises a real-time monitoring module which carries out the monitoring and learning of obtained image data information through a real-time image quality evaluation technology; an alarm system is arranged in the image recognition module, and the alarm system is triggered according to the change of the acquired image data information; the image optimization module receives the alarm signal, locks the damaged position of the image, intercepts image information before and after the image is damaged, and deeply learns the image information by adopting a degraded image restoration algorithm based on a neural network; according to the method, a degraded image restoration algorithm based on a neural network is utilized, image information is deeply learned, clear original information is extracted, fuzzy and de-noising processing is carried out, the damaged position of the image is effectively filled, the monitoring capability is enhanced, direct transformation from the damaged image to the restored image is realized by establishing an end-to-end neural network model, and the image quality is improved. And the image processing efficiency and accuracy are improved.

Description

Technical Field

[0001] The invention relates to an image processing system, in particular to a computer image processing system based on intelligent recognition technology. Background Art

[0002] Computer vision is a branch of computer science that studies how to make computers understand and process image and video data. Its main tasks include image classification, target detection, object recognition, scene understanding, etc. Computer vision requires the support of image processing technology, which provides computer vision with the necessary tools and methods such as image preprocessing, feature extraction, and target detection. The two complement each other and jointly promote the development of computer image processing systems.

[0003] With the continuous development of artificial intelligence technology, digital image processing systems have begun to integrate intelligent algorithms such as deep learning, which can realize intelligent understanding and analysis of image content. This intelligent processing not only improves the accuracy and efficiency of image processing, but also expands the application scenarios of image processing systems. Through training and optimization of large amounts of data, deep learning algorithms can automatically extract useful features from images and classify and identify them. This algorithm optimization enables digital image processing systems to show better robustness and adaptability when facing complex images.

[0004] However, in the prior art, image restoration algorithms are often complex and inefficient, making it difficult to achieve real-time processing, and the restoration effect is limited, making it difficult to fully restore the original clarity and details of the image. In addition, in the prior art, the quality monitoring of image data is often not real-time and accurate enough, and when image quality problems are found, the response speed is slow and cannot be handled in time.

[0005] Therefore, there is an urgent need for a computer image processing system based on intelligent recognition technology to solve the technical problems existing in the above-mentioned prior art. Summary of the invention

[0006] The present invention overcomes the deficiencies of the prior art and provides a computer image processing system based on intelligent recognition technology.

[0007] To achieve the above-mentioned purpose, the technical solution adopted by the present invention is: a computer image processing system based on intelligent recognition technology, comprising: a real-time monitoring module, an image recognition module and an image optimization module; wherein,

[0008] The real-time monitoring module uses real-time image quality assessment technology to monitor and learn the acquired image data information, and improves the monitoring capability of the real-time monitoring module by introducing image signal processing technology and coding algorithm optimization technology; thereby being able to continuously monitor the quality of image data and ensure that the acquired image meets the preset standards.

[0009] An alarm system is provided in the image recognition module. The alarm system is triggered according to changes in the acquired image data information, and an alarm signal is sent to the management personnel and the image optimization module at the same time; thus, it can be triggered immediately when the image data information quality decreases or an abnormality occurs, thereby improving the response speed of the system.

[0010] The image optimization module receives the alarm signal, locks the damaged position of the image, captures the image information before and after the image is damaged, uses a degraded image restoration algorithm based on a neural network, deeply learns the image information, extracts clear original information, blurs and denoises the original information to obtain a degraded image, and then fills the damaged position of the image with the degraded image to enhance the monitoring capability of the real-time monitoring module. In this way, the damaged position of the image can be locked, and the image information before and after the damage can be captured, providing a basis for subsequent image restoration.

[0011] In a preferred embodiment of the present invention, the real-time monitoring module uses an image acquisition device to obtain real-time image data information, and the real-time image quality assessment technology conducts an in-depth analysis of key indicators such as clarity, contrast, color saturation, and noise level in the image data information, and determines whether the quality of the image data meets the requirements by comparing with preset standards or thresholds. The image data is then pre-processed by introducing image signal processing technology to improve the readability and accuracy of the image data. At the same time, the encoding algorithm optimization technology is used to efficiently compress and encode the image data to reduce the demand for storage space and transmission bandwidth.

[0012] In a preferred embodiment of the present invention, the image recognition module checks the changes in the image data information in real time. When it is found that the quality of the image data information has deteriorated or an abnormality has occurred, the image recognition module immediately triggers the alarm system and simultaneously sends an alarm signal to the management personnel and the image optimization module. The management personnel can understand the situation and respond in time, and the image optimization module immediately starts to process the damaged image.

[0013] In a preferred embodiment of the present invention, an end-to-end neural network model is established in the image optimization module, and the image information is used as input, and the degraded image is transformed by the degraded image restoration algorithm to obtain the restored degraded image as output.

[0014] In a preferred embodiment of the present invention, the process of acquiring the degraded image is as follows:

[0015] S1, the neural network obtains image information before and after the image is damaged, and uses the image information as a training data set, and restores the damaged image information through a convolutional neural network or a generative adversarial network;

[0016] S2, training the neural network multiple times using the training data set, reducing the difference between the image information before damage and the image information after damage by using a loss function, and optimizing the neural network;

[0017] S3. Use the optimized neural network to process the damaged image information to obtain the degraded image, input the degraded image into the neural network, and fill the degraded image to the damaged position of the image to obtain a restored image.

[0018] In a preferred embodiment of the present invention, the loss function is expressed as:

[0019]

[0020] In the formula, θ represents the neural network parameters; L() represents the loss function; f θ () represents the neural network model; N represents the number of training data sets; I after Represents the damaged image information; I before Indicates image information before damage.

[0021] In a preferred embodiment of the present invention, the expression formula of the acquisition process of the degraded image is:

[0022] I restored =f θ (I after )

[0023] In the formula, I restored represents the degraded image; f θ () represents the neural network model; I after Indicates damaged image information.

[0024] In a preferred embodiment of the present invention, the expression formula for filling the degraded image to the damaged position of the image is:

[0025]

[0026] In the formula, I filled represents the padded image, that is, the restored image; M is a mask matrix with the same size as the image, which is used to represent the damaged position, that is, the damaged position is 1 and the rest of the positions are 0; Represents element-by-element multiplication; I restored represents the degraded image; I after Indicates damaged image information.

[0027] In a preferred embodiment of the present invention, in step S3), an auxiliary function is used to calculate the image size of the degraded image to adjust the degraded image, and the formula expression of the auxiliary function is:

[0028] I resize = resize(I, target_size)

[0029] Wherein, I represents image data information; resize represents the resizing operation; target_size represents the target size.

[0030] In a preferred embodiment of the present invention, an information data processing terminal is also provided, and the information data processing terminal is used to implement a computer image processing system based on intelligent recognition technology.

[0031] The present invention solves the defects existing in the background technology and has the following beneficial effects:

[0032] (1) The present invention utilizes a neural network-based degraded image restoration algorithm to deeply learn image information, extract clear original information, and perform blurring and denoising processing, thereby effectively filling in the damaged position of the image and enhancing the monitoring capability. In addition, by establishing an end-to-end neural network model, the direct transformation from the damaged image to the restored image is achieved, thereby improving the efficiency and accuracy of image processing.

[0033] (2) The present invention is provided with an alarm system to detect changes in image data in real time, and immediately trigger an alarm when quality problems are found, notifying management personnel and image optimization modules, thereby achieving rapid response and timely processing.

[0034] (3) The present invention can evaluate the quality of image data in real time, and improves the sensitivity and response speed to changes in image data by using image signal processing technology and coding algorithm optimization technology, thereby ensuring the accuracy and readability of the acquired image data information. In addition, by setting an auxiliary function to adjust the image size, the degraded image can be flexibly adapted to different damaged locations, further improving the flexibility and practicality of image restoration. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] 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 are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

[0036] Figure 1 is a system flow chart of a preferred embodiment of the present invention;

[0037] Figure 2 It is a flow chart of the process of acquiring a degraded image according to a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0038] 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.

[0039] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0040] In the description of the present application, it should be understood that the terms "center", "longitudinal", "lateral", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application 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 cannot be understood as limiting the scope of protection of the present application. In addition, the terms "first", "second", etc. are only used for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Therefore, features defined as "first", "second", etc. may explicitly or implicitly include one or more of the features. In the description of the invention, unless otherwise specified, "multiple" means two or more.

[0041] In the description of this application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in this application can be understood by specific circumstances.

[0042] Affected by weather conditions, water, dust and particulate matter may accumulate on the lens of the image acquisition device, all of which will lead to a decrease in image quality. At the same time, strong winds may cause the image acquisition device to shake, thereby causing motion blur in the entire image. In addition, the camera of the image acquisition device used outdoors may cause insects to hit the lens, resulting in blind spots in the camera's field of view, thereby reducing the clarity of the image acquisition and making it difficult to observe. The present invention can reduce the impact of blind spots on the camera content in the first place, thereby improving the clarity of the image.

[0043] Furthermore, the image processing system provided by the present invention can also be applied to a vehicle-mounted system, and when encountering bad weather, it can assist the driver to achieve safe driving and improve driving safety.

[0044] like Figure 1 As shown, a computer image processing system based on intelligent recognition technology includes: a real-time monitoring module, an image recognition module and an image optimization module; wherein,

[0045] The real-time monitoring module uses real-time image quality assessment technology to monitor and learn the acquired image data information. By introducing image signal processing technology and coding algorithm optimization technology, the monitoring capability of the real-time monitoring module is improved. Through real-time image quality assessment technology, the quality of image data can be continuously monitored to ensure that the acquired images meet the preset standards.

[0046] Furthermore, the real-time monitoring module uses image acquisition equipment to obtain real-time image data information. The real-time image quality assessment technology conducts in-depth analysis of key indicators such as clarity, contrast, color saturation, and noise level in the image data information. By comparing with the preset standard or threshold, it is determined whether the quality of the image data meets the requirements. Then, by introducing image signal processing technology, the image data is pre-processed to improve the readability and accuracy of the image data. At the same time, the encoding algorithm optimization technology is used to efficiently compress and encode the image data to reduce the demand for storage space and transmission bandwidth. The introduction of image signal processing technology and encoding algorithm optimization technology improves the readability and accuracy of image data, while reducing the demand for storage space and transmission bandwidth.

[0047] In a preferred embodiment, the image acquisition device can be a video recorder, a monitor, a camera, a sensor, etc.

[0048] An alarm system is set up in the image recognition module. According to the changes in the acquired image data information, the alarm system is triggered and the alarm signal is sent to the management personnel and the image optimization module at the same time. The alarm system is set up so that it can be triggered immediately when the quality of image data information deteriorates or an abnormality occurs, thereby improving the response speed of the system.

[0049] Furthermore, the image recognition module checks the changes in image data information in real time. When it is found that the quality of image data information has deteriorated or an abnormality has occurred, the image recognition module immediately triggers the alarm system and sends an alarm signal to the management personnel and the image optimization module at the same time. The management personnel can understand the situation and respond in time, and the image optimization module immediately starts to process the damaged image.

[0050] The image optimization module receives the alarm signal, locks the damaged position of the image, captures the image information before and after the image is damaged, uses the degraded image restoration algorithm based on neural networks, deeply learns the image information, extracts the clear original information, blurs and denoises the original information to obtain the degraded image, and then fills the damaged position of the image with the degraded image to enhance the monitoring capability of the real-time monitoring module; it can lock the damaged position of the image and capture the image information before and after the damage, which provides a basis for subsequent image restoration. The degraded image restoration algorithm based on neural networks can deeply learn the image information, extract the clear original information, and effectively restore the damaged image.

[0051] Furthermore, an end-to-end neural network model is established in the image optimization module, which takes image information as input and transforms it through a degraded image restoration algorithm to obtain a restored degraded image as output.

[0052] like Figure 2 As shown in Figure 2, the process of obtaining the degraded image is as follows:

[0053] S1. The neural network obtains the image information before and after the image is damaged, and uses it as a training data set to restore the damaged image information through a convolutional neural network or a generative adversarial network;

[0054] S2, training the neural network multiple times using the training data set, reducing the difference between the image information before damage and the image information after damage through the loss function, and optimizing the neural network;

[0055] Furthermore, the loss function is expressed as:

[0056]

[0057] In the formula, θ represents the neural network parameters; L() represents the loss function; f θ () represents the neural network model; N represents the number of training data sets; I after Represents the damaged image information; I before Indicates image information before damage.

[0058] S3. Use the optimized neural network to process the damaged image information to obtain a degraded image, input the degraded image into the neural network, and fill the degraded image to the damaged position of the image to obtain a restored image.

[0059] The expression formula of the acquisition process of the degraded image is:

[0060] I restored =f θ (I after )

[0061] In the formula, I restored represents the degraded image; f θ () represents the neural network model; I after Represents the damaged image information. The image information before and after the damage is restored through the neural network, which improves the accuracy of image restoration. The loss function is used to reduce the difference between the image information before and after the damage, optimize the neural network model, and further improve the effect of image restoration.

[0062] The expression formula for filling the degraded image to the damaged position of the image is:

[0063]

[0064] In the formula, I filled represents the padded image, that is, the restored image; M is a mask matrix with the same size as the image, which is used to represent the damaged position, that is, the damaged position is 1 and the rest of the positions are 0; Represents element-by-element multiplication; I restored represents the degraded image; I after Represents the damaged image information. The damaged position is accurately located through the mask matrix, and the degraded image is filled to that position, thus achieving complete image restoration.

[0065] In step S3), an auxiliary function is used to calculate the image size of the degraded image to adjust the degraded image. The formula expression of the auxiliary function is:

[0066] I resize = resize(I, target_size)

[0067] In the formula, I represents image data information; resize represents the resizing operation; target_size represents the target size. The auxiliary function can calculate and adjust the size of the degraded image to ensure that the padded image is the same size as the original image, thus improving the overall coordination of the image.

[0068] Embodiment 1

[0069] In the fields of intelligent monitoring, self-driving cars, etc., cameras are important perception devices, and their image quality directly affects the performance and safety of the system. However, in practical applications, camera lenses are easily affected by factors such as weather and environment, resulting in a decrease in image quality. In particular, emergencies such as insects hitting the screen will leave stains on the lens, affecting the clarity of subsequent videos. In order to solve this problem, this experiment will apply a system based on real-time monitoring and image restoration technology to automatically detect and repair stains on camera lenses to improve image clarity and system robustness.

[0070] This embodiment will simulate the situation in which the lens of an intelligent monitoring device is accidentally hit by an insect and produces stains during outdoor use, and then optimize and repair the image of the portion of the monitoring video where the stains are produced.

[0071] The following are the related equipment used in this embodiment:

[0072] High-definition camera: with a resolution of 1920x1080 and a frame rate of 30fps, used to shoot experimental videos.

[0073] Control system: Based on PC, equipped with special image processing software to control camera shooting and real-time monitoring and processing of images.

[0074] Insect impact simulation device: driven by a small motor, it can simulate insect impacts at different speeds and angles.

[0075] Stain Making Materials: Use common inks whose colors and textures are similar to common insect residues.

[0076] Computer and image processing software: The computer configuration is Intel i7 processor, 16GB memory, NVIDIA GTX1080 graphics card. The image processing software is a self-developed real-time image monitoring and repair system.

[0077] Experimental preparation: Fix the high-definition camera on the experimental table, adjust the shooting angle and focal length to ensure that the captured image is clear and stable, and set the focal length to 50mm. Connect the camera to the control system, set the camera shooting parameters, resolution 1080P, frame rate 30 frames / second, and ensure that the camera can shoot according to the preset parameters, prepare the simulated insect impact device and stain making materials, and set the speed and angle parameters of the impact device to ensure the authenticity and diversity of the simulated insect impact.

[0078] Start the camera and start shooting a 5-minute clear image video as the control group. The shooting environment is indoors with stable lighting conditions to avoid interference from external factors.

[0079] After the camera has been filming for 2 minutes, use the simulated insect impact device to simulate the situation where an insect hits the screen. Set the impact speed to 2m / s and the impact angle to 45 degrees.

[0080] Use stain making material to make a stain with a diameter of 2 mm at the impact position. Make sure that the color and texture of the stain are similar to the insect residue in actual use. Continue to shoot the video to record the impact of the stain on the image quality for 3 minutes.

[0081] Import the captured video into the computer and use image processing software for real-time monitoring and processing. Set the spot detection threshold as follows: when the contrast difference between the spot and the surrounding environment exceeds 10%; when the brightness value of a certain pixel is lower or higher than 15% of the average brightness value of the surrounding pixels; the real-time monitoring module monitors the video stream in real time to ensure that any changes in image quality can be captured.

[0082] During the experiment, once the image data information quality deteriorates or becomes abnormal, the image recognition module will immediately trigger the alarm system and send an alarm signal to the management personnel and the image optimization module at the same time.

[0083] After receiving the alarm signal from the image recognition module, the image optimization module immediately starts to process the damaged image. It locks the damaged position of the image and intercepts the image information before and after the image is damaged. The neural network model in the image optimization module restores the damaged image information through a convolutional neural network or a generative adversarial network. The neural network is trained multiple times using the training data set, and the difference between the image information before and after the damage is reduced through the loss function to optimize the neural network. Finally, the damaged image information is processed using the optimized neural network to obtain a degraded image, which is then filled to the damaged position of the image to obtain a restored image.

[0084] When the image optimization module processes a damaged image, the loss function expression formula used is used to measure the effect of the neural network in restoring the image. It represents the relationship between the neural network parameters, the loss function, the neural network model, the number of training data sets, the image information after damage, and the image information before damage. By minimizing the loss function, we can optimize the parameters of the neural network to better restore the damaged image.

[0085] The expression formula of the degraded image acquisition process describes how to use the neural network model and the damaged image information to generate the degraded image. This formula is the core part of the image optimization module, which realizes the conversion from damaged image to degraded image.

[0086] The expression formula for filling the degraded image to the damaged position of the image describes how to use the degraded image to repair the damaged image. This formula combines the degraded image with the damaged position by element-by-element multiplication to obtain the restored image.

[0087] The auxiliary function formula expression is used to resize the degraded image to ensure that it can accurately fill the damaged location.

[0088] In summary, through the collaborative work of the real-time monitoring module, image recognition module and image optimization module, our system can monitor and repair the stain problem on the camera lens in real time, improve the clarity of the image and the robustness of the system. During the experiment, we verified the feasibility and effectiveness of the system, and recorded the key parameters and repair effects for subsequent analysis.

[0089] Embodiment 2

[0090] Reduced visibility and insufficient brightness in foggy days pose a serious threat to driver safety, increasing the risk of vehicle collisions and being the main cause of casualties on foggy roads. Fog droplets suspended in the fog block and scatter light, which results in less light reaching the driver's eyes, lower contrast, and therefore reduced visibility.

[0091] Therefore, many camera-based advanced driver assistance systems (ADAS) have been introduced to assist drivers and ensure their safety in various driving conditions.

[0092] In this embodiment, the technical solution proposed by the present invention is introduced into an advanced driver assistance system (ADAS). When the system detects foggy conditions, it uses advanced image processing algorithms to enhance the real-time video stream captured by the camera. By improving the contrast and brightness of the image, the light scattering and blocking effects caused by fog droplets are reduced, thereby significantly improving the driver's vision.

[0093] The system will continuously analyze the processed image data and dynamically evaluate the visibility conditions of the current road. This information will be used to adjust the vehicle's safe driving strategy, such as automatically slowing down or increasing the following distance. Based on the enhanced image and visibility assessment, if the system detects a potential dangerous situation, such as a sudden deceleration of the vehicle ahead or the appearance of a pedestrian, it will immediately send a warning signal to the driver. If necessary, the system can also take auxiliary driving measures, such as automatic braking or obstacle avoidance steering, to prevent a collision.

[0094] Through the implementation of the above technical solutions, the present invention can significantly improve the driver's safety and driving experience in foggy conditions and reduce the risk of vehicle collision due to reduced visibility.

[0095] When the present invention is used, the system first uses the image acquisition device to obtain real-time image data information through the real-time monitoring module, and the real-time image quality evaluation technology performs in-depth analysis on the obtained image data, including key indicators such as clarity, contrast, color saturation, noise level, etc., and compares it with preset standards or thresholds to determine whether the quality of the image data meets the requirements.

[0096] If the image quality does not meet the requirements, the system will introduce image signal processing technology to pre-process the image data to improve the readability and accuracy of the image data. At the same time, the system will use encoding algorithm optimization technology to efficiently compress and encode the image data to reduce the demand for storage space and transmission bandwidth.

[0097] The image recognition module will check the changes in image data information in real time. Once it finds that the quality of image data information has deteriorated or an abnormality has occurred, the image recognition module will immediately trigger the alarm system and send an alarm signal to the management personnel and the image optimization module at the same time.

[0098] After receiving the alarm signal, the image optimization module will immediately start processing the damaged image. The module will lock the damaged position of the image and intercept the image information before and after the damage. Using the degraded image restoration algorithm based on neural network, deep learning is performed on the damaged image to extract clear original information. After blurring and denoising, the degraded image is obtained and filled to the damaged position of the image, thereby enhancing the monitoring capability of the real-time monitoring module.

[0099] The optimized images are output and displayed for further analysis and processing by managers or related systems. Managers can provide feedback to the system based on the restored images to further adjust and optimize the system. The system will also conduct self-learning and iterative optimization based on historical data and feedback information to improve future image processing effects and accuracy.

[0100] The above is based on the ideal embodiment of the present invention. Through the above description, relevant personnel can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the content in the specification, and the technical scope must be determined according to the scope of the claims.

Claims

1. A computer image processing system based on intelligent recognition technology, characterized in that: include: Real-time monitoring module, image recognition module and image optimization module; among them, The real-time monitoring module uses real-time image quality assessment technology to monitor and learn the acquired image data information, and improves the monitoring capability of the real-time monitoring module by introducing image signal processing technology and coding algorithm optimization technology; The image recognition module is provided with an alarm system, which is triggered according to changes in the acquired image data information, and sends an alarm signal to the management personnel and the image optimization module at the same time; The image optimization module receives the alarm signal, locks the damaged position of the image, captures the image information before and after the image is damaged, adopts a degraded image restoration algorithm based on a neural network, deeply learns the image information, extracts clear original information, blurs and denoises the original information to obtain a degraded image, and then fills the damaged position of the image with the degraded image to enhance the monitoring capability of the real-time monitoring module.

2. A computer image processing system based on intelligent recognition technology according to claim 1, characterized in that: The real-time monitoring module uses an image acquisition device to obtain real-time image data information. The real-time image quality assessment technology conducts in-depth analysis on key indicators such as clarity, contrast, color saturation, and noise level in the image data information, and determines whether the quality of the image data meets the requirements by comparing with preset standards or thresholds. The image data is then preprocessed by introducing image signal processing technology to improve the readability and accuracy of the image data. At the same time, the encoding algorithm optimization technology is used to efficiently compress and encode the image data to reduce the demand for storage space and transmission bandwidth.

3. The computer image processing system based on intelligent recognition technology according to claim 1, characterized in that: The image recognition module checks the changes of the image data information in real time. When it is found that the quality of the image data information has deteriorated or an abnormality has occurred, the image recognition module immediately triggers the alarm system and simultaneously sends an alarm signal to the management personnel and the image optimization module. The management personnel can understand the situation in time and respond, and the image optimization module immediately starts to process the damaged image.

4. The computer image processing system based on intelligent recognition technology according to claim 1, characterized in that: An end-to-end neural network model is established in the image optimization module, which takes the image information as input and transforms it through the degraded image restoration algorithm to obtain the restored degraded image as output.

5. The computer image processing system based on intelligent recognition technology according to claim 4, characterized in that: The process of obtaining the degraded image is as follows: S1, the neural network obtains image information before and after the image is damaged, and uses the image information as a training data set, and restores the damaged image information through a convolutional neural network or a generative adversarial network; S2, training the neural network multiple times using the training data set, reducing the difference between the image information before damage and the image information after damage by using a loss function, and optimizing the neural network; S3. Use the optimized neural network to process the damaged image information to obtain the degraded image, input the degraded image into the neural network, and fill the degraded image to the damaged position of the image to obtain a restored image.

6. The computer image processing system based on intelligent recognition technology according to claim 5, characterized in that: The loss function is expressed as: In the formula, θ represents the neural network parameters; L() represents the loss function; f θ () represents the neural network model; N represents the number of training data sets; I after Represents the damaged image information; I before Indicates image information before damage.

7. A computer image processing system based on intelligent recognition technology according to claim 5, Features: The expression formula of the acquisition process of the degraded image is: I restored =f θ (Ia fter ) In the formula, I restored represents the degraded image; f θ () represents the neural network model; I after Indicates damaged image information.

8. The computer image processing system based on intelligent recognition technology according to claim 5, characterized in that: The expression formula for filling the degraded image to the damaged position of the image is: In the formula, I filled represents the padded image, that is, the restored image; M is a mask matrix with the same size as the image, which is used to represent the damaged position, that is, the damaged position is 1 and the rest of the positions are 0; Represents element-by-element multiplication; I restored represents the degraded image; I after Indicates damaged image information.

9. The computer image processing system based on intelligent recognition technology according to claim 5, characterized in that: In step S3), an auxiliary function is used to calculate the image size of the degraded image to adjust the degraded image. The formula expression of the auxiliary function is: I resize =resize(I,target_size) Wherein, I represents image data information; resize represents the resizing operation; target_size represents the target size.

10. An information data processing terminal, characterized in that: The information data processing terminal is used to implement a computer image processing system based on intelligent recognition technology as described in any one of claims 1-9.

Images