Machine vision unit applied to intelligent security management system of intelligent community

Abstract

The invention discloses a machine vision unit applied to a smart community intelligent security management system, and relates to the technical field of image processing, and the machine vision unit comprises an image obtaining module which obtains a real-time video stream monitored at night, and carries out the frame cutting to obtain a plurality of frames of images; the multi-frame image fusion module is used for processing the multi-frame image by using a multi-exposure fusion algorithm to generate a fused image; the self-adaptive correction module is used for self-adaptively adjusting the brightness and the contrast ratio of the fused image and carrying out detail enhancement and color correction; the image processing module is used for removing noise and enhancing details through bilateral filtering and a multi-scale Retinex algorithm; and the image post-processing module is used for performing color balance adjustment to ensure that the images are natural and consistent. Through multi-frame image fusion, adaptive correction, multi-scale detail enhancement and color balance processing, the definition, detail expression and recognition precision of the image in a night monitoring scene are remarkably improved, the adaptability and robustness of the system are enhanced, and meanwhile redundant information and resource consumption in the processing process are reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and in particular to a machine vision unit applied to an intelligent security management system of a smart community. Background Art

[0002] With the widespread application of smart communities and intelligent security systems, traditional video surveillance technology can no longer meet the needs of modern, intelligent security systems. In existing technologies, during nighttime monitoring, light pollution (such as strong light from street lamps, car headlights, and advertising screens) can interfere with the image quality of the camera, often resulting in problems such as image quality degradation, color distortion, and loss of details, which affects the monitoring effect and the accuracy of event recognition.

[0003] At present, commonly used image enhancement and correction methods include histogram equalization, filter enhancement, Retinex algorithm, etc. These methods can improve image quality to a certain extent, but usually have defects such as unstable processing effect, high computational complexity, and poor real-time performance. In addition, the existing multi-image fusion technology is prone to information loss or redundancy when processing different viewing angles and multi-source information of surveillance videos, making it difficult to meet the efficiency and accuracy requirements of real-time monitoring. Summary of the invention

[0004] Based on the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide a machine vision unit applied to a smart community intelligent security management system to solve the above-mentioned technical problems.

[0005] To achieve the above object, the present invention provides the following technical solution: a machine vision unit applied to a smart community intelligent security management system, comprising:

[0006] Image acquisition module: obtains the real-time video stream of all security monitoring at night in the smart community intelligent security management system, performs frame capture processing on the real-time video stream and classifies it by scene to obtain multiple frames of images;

[0007] Multi-frame image fusion module: Through the multi-exposure fusion algorithm, multiple frames of images in the same scene are fused to obtain a fused image;

[0008] Adaptive correction module: Adaptive gamma correction algorithm and multi-channel image processing method are used to adaptively adjust the brightness and contrast of the fused image according to the local features of the fused image, and multi-channel detail enhancement and color correction are performed on the fused image after brightness and contrast adjustment to obtain a corrected image;

[0009] Image processing module: The corrected image is processed using a bilateral filtering algorithm and a multi-scale Retinex algorithm, wherein the bilateral filtering algorithm is used to remove high-frequency noise in the corrected image, and the multi-scale Retinex algorithm is used to perform nonlinear adjustment on the brightness of the corrected image to enhance the detail performance of the corrected image;

[0010] Image post-processing module: post-process and adjust the color balance of the processed corrected image to ensure the natural color and visual consistency of the image.

[0011] The present invention is further configured such that the multi-frame image fusion module includes a feature weight calculation component and a multi-exposure fusion component; the feature weight calculation component is used to preprocess multi-frame images taken by a camera at different exposure times in the same scene, and calculate the feature weight of each image; the multi-exposure fusion component is used to fuse the multi-frame images through a multi-exposure fusion algorithm according to the feature weight.

[0012] The present invention is further configured such that the calculation logic of the feature weight of each image is: Among them, W k (x, y) is the feature weight of the k-th frame image at the pixel position (x, y), I k (x, y) is the pixel value of the k-th frame image at the pixel position (x, y), M is the median brightness of the image, σ is the brightness expansion parameter, is the image gradient, which is used to measure the degree of change of image pixel values, and α is the gradient enhancement coefficient;

[0013] According to the feature weights, multiple frames of images are fused through a multi-exposure fusion algorithm, and the fusion logic is: Among them, I fusion (x, y) is the pixel value of the fused image at the pixel position (x, y), N is the number of multi-frame images, and β is the nonlinear enhancement coefficient, which is used to control the nonlinear effect of fusion.

[0014] The present invention is further configured such that the adaptive correction module includes a gamma correction component and a multi-channel correction component; the gamma correction component is used to adaptively adjust the brightness and contrast of the fused image according to the local features of the fused image by using an adaptive gamma correction algorithm; the multi-channel correction component is used to perform multi-channel detail enhancement and color correction on the fused image after the brightness and contrast are adjusted by using a multi-channel image processing method to obtain a corrected image.

[0015] The present invention is further configured such that the calculation logic of the adaptive gamma correction algorithm is: adjust (x,y)=I fusion (x,y) γ(x,y) , where Iadjust (x, y) is the pixel value of the image at the pixel position (x, y) after brightness and contrast adjustment. fusion (x, y) is the pixel value of the fused image at the pixel position (x, y), γ(x, y) is the adaptive gamma coefficient, and the calculation logic of the adaptive gamma coefficient γ(x, y) is: Among them, η is the contrast enhancement strength parameter, κ is the brightness adjustment sensitivity parameter, and T is the local brightness threshold, which is used to control the benchmark of gamma adjustment.

[0016] The present invention is further configured to perform multi-channel detail enhancement and color correction on the fused image after brightness and contrast adjustment, and the calculation logic is: Among them, I checkl (x, y) is the pixel value of the corrected image at the pixel position (x, y), I adjust,c (x, y) is the pixel value of the image of the cth channel of the adjusted multi-channel image at the pixel position (x, y), is the image gradient, which is used to measure the degree of change of the image pixel value, τ is a nonlinear parameter that controls the degree of detail enhancement, and c∈{R,G,B} indicates that the R, G and B color channels of the image are calculated separately.

[0017] The present invention is further configured such that the image processing module includes a bilateral filtering denoising component and a multi-scale Retinex algorithm enhancement component; the bilateral filtering denoising component is used to perform bilateral filtering on the corrected image, retain edge features in the corrected image, and remove high-frequency noise in the corrected image; the multi-scale Retinex algorithm enhancement component is used to perform nonlinear adjustment on the brightness of the corrected image, and enhance the detail performance of the corrected image.

[0018] The present invention is further configured such that the calculation logic of bilateral filtering of the corrected image is: Among them, I bf (x, y is the pixel value of the corrected image at the pixel position (x, y) after bilateral filtering, Ω is the local window range, ||I checkl (u,v)-I checkl (x,y)|| is the difference between the pixel values ​​at the pixel position (u,v) and the pixel value at the pixel position (x,y), ||(u,vA-(x,y)|| is the spatial distance between the pixel position (u,v) and the pixel position (x,y), σ r and σ s are the control parameters of pixel value difference and spatial distance respectively, and FC(x,y) is the normalization factor.

[0019] The present invention is further configured such that the calculation logic for nonlinearly adjusting the brightness of the corrected image is: Among them, I msr (x, y) is the pixel value of the corrected image at the pixel position (x, y) after multi-scale Retinex algorithm enhancement, L is the number of scales, ω n is the weight coefficient of the nth scale, The scale is σ n Gaussian convolution kernel, * represents convolution operation.

[0020] The present invention is further configured to perform post-processing and color balance adjustment on the processed corrected image, and the calculation logic is: Among them, I final (x, y) is the pixel value of the final optimized image at the pixel position (x, y), λ c Add weight to the colors, is the color gradient, which is used to measure the degree of color change in an image. is the image gradient, which is used to measure the degree of change of the image pixel value, and φ is the color balance control parameter, which is used to adjust the intensity of color balance.

[0021] The present invention provides a machine vision unit applied to a smart community intelligent security management system, comprising an image acquisition module: acquiring real-time video streams of all security monitoring at night in the smart community intelligent security management system, performing frame capture processing on the real-time video streams and classifying them according to scenes to obtain multiple frames of images; a multi-frame image fusion module: fusing multiple frames of images in the same scene through a multi-exposure fusion algorithm to obtain a fused image; an adaptive correction module: using an adaptive gamma correction algorithm and a multi-channel image processing method, adaptively adjusting the brightness and contrast of the fused image according to the local features of the fused image, and performing multi-channel detail enhancement and color correction on the fused image after the brightness and contrast are adjusted to obtain a corrected image; an image processing module: using a bilateral filtering algorithm and a multi-scale Retinex algorithm to process the corrected image, wherein the bilateral filtering algorithm is used to remove high-frequency noise in the corrected image, and the multi-scale Retinex algorithm is used to perform nonlinear adjustment on the brightness of the corrected image to enhance the detail performance of the corrected image; an image post-processing module: performing post-processing and color balance adjustment on the processed corrected image to ensure the natural color and visual consistency of the image, and the beneficial effects produced include:

[0022] 1. Improve the clarity and detail of nighttime surveillance images: Through the multi-frame image fusion module, the multi-exposure fusion algorithm is used to fuse multiple frames of images in the same scene, effectively enhancing the dynamic range of the image, so that the details of the highlight and dark areas are fully preserved and displayed; the fused image can better restore the real scene, avoid overexposure or underexposure, and improve the overall clarity and detail of nighttime surveillance images;

[0023] 2. Enhance image quality in light-polluted scenes: Bilateral filtering and multi-scale Retinex algorithms are used to process the corrected image. Bilateral filtering effectively removes high-frequency noise in the image while maintaining the sharpness of edges and details. The multi-scale Retinex algorithm further performs nonlinear brightness adjustment to enhance image detail performance, so that target objects in the monitoring screen can be clearly identified in light-polluted environments, improving the image quality and recognition accuracy of the intelligent security management system;

[0024] 3. Implement adaptive correction to improve the adaptability of the system: The adaptive correction module combines adaptive gamma correction and multi-channel image processing methods to automatically adjust the brightness and contrast of the fused image according to the local features of the image, and implement real-time adaptive correction for different lighting conditions. This adaptive mechanism enhances the robustness and adaptability of the system in various lighting environments, helping to reduce false positives and false negatives;

[0025] 4. Maintain the natural color and visual consistency of the image: The image post-processing module ensures the consistency of the natural color and visual effect of the final image through color balance adjustment. By performing multi-channel color enhancement and correction on the processed corrected image, the color cast and distortion problems are avoided, making the monitoring picture more realistic and natural, improving the visual experience of the image and the user-friendliness of the system;

[0026] 5. Reduce redundant information and improve processing efficiency: By classifying and optimizing the acquired multi-frame images, it can effectively reduce redundant data, reduce unnecessary computing burden, and improve image processing efficiency. Combined with the optimization algorithms of each module, it effectively reduces the occupation of system resources and enhances the real-time performance and response speed of the intelligent security management system.

[0027] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. In the drawings:

[0029] Figure 1 The present invention is a schematic diagram showing the structure of a machine vision unit applied to an intelligent security management system for a smart community, showing an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0030] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0031] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0032] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it is obvious to those skilled in the art that the embodiments of the present invention can be implemented without these specific details. In other embodiments, well-known structures and devices are shown in the form of block diagrams rather than in detail to avoid making the embodiments of the present invention difficult to understand.

[0033] A machine vision unit used in smart community intelligent security management system, such as Figure 1 As shown, including:

[0034] Image acquisition module: obtains the real-time video stream of all security monitoring at night in the smart community intelligent security management system, performs frame capture processing on the real-time video stream and classifies it by scene to obtain multiple frames of images;

[0035] Multi-frame image fusion module: Through the multi-exposure fusion algorithm, multiple frames of images in the same scene are fused to obtain a fused image;

[0036] Adaptive correction module: Adaptive gamma correction algorithm and multi-channel image processing method are used to adaptively adjust the brightness and contrast of the fused image according to the local features of the fused image, and multi-channel detail enhancement and color correction are performed on the fused image after brightness and contrast adjustment to obtain a corrected image;

[0037] Image processing module: The corrected image is processed using a bilateral filtering algorithm and a multi-scale Retinex algorithm, wherein the bilateral filtering algorithm is used to remove high-frequency noise in the corrected image, and the multi-scale Retinex algorithm is used to perform nonlinear adjustment on the brightness of the corrected image to enhance the detail performance of the corrected image;

[0038] Image post-processing module: post-process and adjust the color balance of the processed corrected image to ensure the natural color and visual consistency of the image.

[0039] Specifically, video stream data is obtained in real time from each surveillance camera deployed in the community. The camera works continuously to capture the surveillance scene at night. The video stream data format is a continuous frame sequence, usually transmitted at a frame rate. Real-time video stream data provides continuous scene information to ensure full coverage and continuous monitoring of the entire surveillance area; regular frame capture is performed from the real-time video stream to extract representative frame images. Frame capture processing includes: frame sampling: based on the dynamic characteristics of the scene, the video stream is sampled at time intervals; key frame selection: further select frames with high information content as key frames, filter out redundant information, and reduce the complexity of data storage and processing. Set according to actual needs, and there is no restriction here.

[0040] The present invention is further configured such that a multi-frame image fusion module fuses multiple frames of images in the same scene through a multi-exposure fusion algorithm to obtain a fused image; specifically, the multi-frame image fusion module includes a feature weight calculation component and a multi-exposure fusion component; the feature weight calculation component is used to pre-process multiple frames of images taken by a camera at different exposure times in the same scene, and calculate the feature weight of each image; the multi-exposure fusion component is used to fuse multiple frames of images through a multi-exposure fusion algorithm according to the feature weight; specifically, the feature weight calculation component is responsible for calculating the feature weight of each frame of the image to determine the contribution of each frame of the image in the fusion process; the multi-exposure fusion component uses the calculated feature weights to perform weighted fusion processing on multiple frames of images through a multi-exposure fusion algorithm.

[0041] The present invention is further configured such that the calculation logic of the feature weight of each image is: Among them, W k (x, y) is the feature weight of the k-th frame image at the pixel position (x, y), I k (x, y) is the pixel value of the k-th frame image at the pixel position (x, y), M is the median brightness of the image, σ is the brightness expansion parameter, is the image gradient, which is used to measure the degree of change of the image pixel value, and α is the gradient enhancement coefficient; specifically, the above calculation logic is used to determine the feature weight W of each pixel in each image during the multi-frame fusion process.k (x,y). By quantifying the specific features of the image, combined with factors such as brightness, median deviation, gradient change, etc., the importance of each pixel in the fusion process is calculated. Specifically, the calculation of weights is divided into two parts: the weight of brightness difference and the weight of gradient enhancement. Calculate the difference between the pixel brightness value and the median brightness of the image as a whole, and convert the brightness difference into a weight. The smaller the difference, the greater the weight. The parameter σ adjusts the sensitivity of the weight to the brightness difference. Smaller differences will get higher weights. The gradient of the image is calculated for weight adjustment, which indicates the amplitude of the pixel value change in the image. A larger gradient usually means an edge or detail area. The parameter α is used to enhance the sensitivity to gradient changes. Furthermore, the brightness expansion parameter σ controls the influence range of brightness difference on weight calculation. It is selected according to the brightness distribution characteristics of the image, and the value range is 1 / 4 to 1 / 2 of the brightness value. For example, for 8-bit images, it is usually [10,50]. The gradient enhancement coefficient α controls the sensitivity of gradient enhancement, which is used to amplify or reduce the influence of gradient changes on weight calculation, and the value range is [1,5]. By reasonably calculating the feature weight of each pixel, important brightness information and edge details in the image are retained, and the quality of image fusion is effectively improved, so that the details of the bright and dark areas of the final output image are clearer. .

[0042] According to the feature weights, multiple frames of images are fused through a multi-exposure fusion algorithm, and the fusion logic is: Among them, I fusion (x, y) is the pixel value of the fused image at the pixel position (x, y), N is the number of multi-frame images, and β is the nonlinear enhancement coefficient, which is used to control the nonlinear effect of fusion; the above calculation logic generates the final fused image by weighted fusion of multiple frames of images in the same scene. The fusion process is based on the feature weights of each frame image. The nonlinear enhancement coefficient β is used to perform weighted averaging of the pixel values ​​at each pixel position to ensure that the key information of each frame image is effectively retained during the fusion process, thereby improving the dynamic range and visual quality of the final output image. The nonlinear enhancement coefficient β adjusts the enhancement effect of brightness and contrast during the fusion process. Higher values ​​increase the brightness and contrast of the image. The value range is [0.5, 2.0]. Values ​​below 1 are used to reduce brightness, and values ​​above 1 are used to enhance brightness. The multi-exposure fusion algorithm effectively retains the details of bright and dark parts by integrating multi-frame image information, increases the dynamic range of the image, and enables the fused image to be better presented under complex lighting conditions.

[0043] The present invention is further configured such that the adaptive correction module includes a gamma correction component and a multi-channel correction component; the gamma correction component is used to adaptively adjust the brightness and contrast of the fused image according to the local features of the fused image by using an adaptive gamma correction algorithm; the multi-channel correction component is used to perform multi-channel detail enhancement and color correction on the fused image after the brightness and contrast are adjusted by using a multi-channel image processing method to obtain a corrected image. Specifically, the gamma correction component uses an adaptive gamma correction algorithm to dynamically adjust the gamma value of the image according to the local features of the fused image, thereby adaptively optimizing the overall brightness and contrast of the image. The multi-channel correction component further processes the gamma-corrected fused image, and uses a multi-channel image processing method to perform independent detail enhancement and color correction on different color channels (including red, green and blue channels), ensuring that the color of the image remains natural and consistent, while improving the detail expression of the image in different channels.

[0044] The present invention is further configured such that the calculation logic of the adaptive gamma correction algorithm is: adjust (x,y)=I fusion (x,y) γ(x,y) , where I adjust (x, y) is the pixel value of the image at the pixel position (x, y) after brightness and contrast adjustment. fusion (x, y) is the pixel value of the fused image at the pixel position (x, y), γ(x, y) is the adaptive gamma coefficient, and the calculation logic of the adaptive gamma coefficient γ(x, y) is: Among them, η is the contrast enhancement strength parameter, κ is the brightness adjustment sensitivity parameter, and T is the local brightness threshold, which is used to control the benchmark of gamma adjustment. Specifically, the above calculation logic dynamically adjusts the gamma coefficient according to the pixel value of the fused image, and performs differentiated brightness correction for image areas of different brightness, thereby optimizing the brightness and contrast performance of the overall image; the contrast enhancement strength parameter η adjusts the gain of the gamma correction to enhance the image contrast in the range of [0.5,3.0]. A higher value will result in a stronger contrast enhancement, while a lower value will maintain a smoother brightness adjustment; the brightness adjustment sensitivity parameter κ controls the sensitivity of the gamma coefficient to brightness changes. A larger value makes the gamma adjustment more sensitive to brightness changes, while a smaller value makes the adjustment smoother; the value range is [0.1,10]. When it is lower than 1, the adjustment is softer, and when it is higher than 1, the adjustment is more aggressive; the local brightness threshold T determines the reference brightness for gamma adjustment, which is usually selected as the average brightness of the image or a predetermined brightness reference value. For 8-bit images, the value range is [0,255]; the adaptive gamma correction method can dynamically adjust the gamma coefficient according to the local brightness characteristics of the image, optimize the visual effects of bright and dark areas, and make the image more detailed under various lighting conditions.

[0045] The present invention is further configured to perform multi-channel detail enhancement and color correction on the fused image after brightness and contrast adjustment, and the calculation logic is: Among them, I checkl (x, y) is the pixel value of the corrected image at the pixel position (x, y), I adjust,c (x, y) is the pixel value of the image of the cth channel of the adjusted multi-channel image at the pixel position (x, y), is the image gradient, which is used to measure the degree of change in the image pixel value. τ is a nonlinear parameter that controls the degree of detail enhancement. c∈{R,G,B} indicates that the R, G, and B color channels of the image are calculated separately. The above calculation logic further optimizes the visual effect and color performance of the image by performing multi-channel (including R, G, and B channels) detail enhancement and color correction on the fused image after brightness and contrast adjustment. Each channel is processed separately to ensure that the details and colors of different color channels can be properly adjusted and optimized. The nonlinear parameter τ that controls the degree of detail enhancement controls the intensity of detail enhancement. Larger values ​​are used for stronger detail enhancement, while smaller values ​​are used to maintain more natural detail performance. The value range is [0.5,5.0]. By performing nonlinear enhancement on the image gradient of each color channel, the detail performance of the image can be significantly improved, especially in the edge and texture-rich areas, and the visual clarity of the image can be enhanced.

[0046] The present invention is further configured as follows: the image processing module includes a bilateral filtering denoising component and a multi-scale Retinex algorithm enhancement component; the bilateral filtering denoising component is used to perform bilateral filtering on the corrected image, retain edge features in the corrected image, and remove high-frequency noise in the corrected image; the multi-scale Retinex algorithm enhancement component is used to perform nonlinear adjustment on the brightness of the corrected image to enhance the detail performance of the corrected image; specifically, the bilateral filtering denoising component uses bilateral filtering technology to perform denoising on the corrected image. The bilateral filtering technology can retain edge information while removing high-frequency noise in the image, and allocates weights by considering the similarity of pixel values ​​and pixel positions, thereby avoiding the edge blurring problem caused by traditional linear filters in the denoising process; the multi-scale Retinex algorithm enhancement component is used to perform nonlinear adjustment on the brightness of the corrected image to enhance the detail performance of the image. The Retinex algorithm simulates the characteristics of the human visual system and improves the visual effect of the image by balancing brightness at different scales.

[0047] The present invention is further configured such that the calculation logic of bilateral filtering of the corrected image is: Among them, I bf(x, y) is the pixel value of the corrected image at the pixel position (x, y) after bilateral filtering, Ω is the local window range, ||I checkl (u,v)-I checkl (x,y)|| is the difference between the pixel values ​​at the pixel position (u,v) and the pixel value at the pixel position (x,y), ||(u,v)-(x,y)|| is the spatial distance between the pixel position (u,v) and the pixel position (x,y), σ r and σ s are the control parameters of pixel value difference and spatial distance respectively, and FC(x,y) is the normalization factor. In the above calculation logic, the weighted average value of each pixel is determined according to the relative position between different pixels and the similarity of pixel values ​​to obtain a smooth and clear image. The pixel value difference control parameter σ r Determines the influence of pixel value similarity on weights during filtering. A smaller value means that only pixels of similar colors will affect each other, thereby better preserving edges; a larger value allows more pixels to participate in weighted averaging, resulting in a stronger denoising effect but may lose edge information. The value range is [10,50] and is adjusted according to the noise level of the image and the edge features that need to be preserved; the spatial distance control parameter σ s Controls the effect of pixel position distance on weight during filtering. A smaller value means that only nearby pixels will affect the filtering result; a larger value allows farther pixels to affect the current pixel. The value range is [1,10], which is selected according to the image resolution and neighborhood size. The calculation logic of the normalization factor FC(x,y) is: It is used to ensure that the sum of the weighted average is 1, thereby keeping the pixel value within a reasonable range. The normalization factor is the sum of all weights, ensuring the brightness consistency of the filtering result with the original image. Bilateral filtering combines the weight calculation of the spatial and pixel value domains to effectively remove high-frequency noise from the image while retaining the edge and detail information of the image, thereby improving the visual quality of the image.

[0048] The present invention is further configured such that the calculation logic for nonlinearly adjusting the brightness of the corrected image is: Among them, I msr (x, y) is the pixel value of the corrected image at the pixel position (x, y) after multi-scale Retinex algorithm enhancement, L is the number of scales, ω n is the weight coefficient of the nth scale, The scale is σ nThe Gaussian convolution kernel is represented by *, which indicates the convolution operation. Specifically, the above calculation logic uses the Multi-Scale Retinex (MSR) algorithm to perform nonlinear adjustment on the brightness of the corrected image to enhance the image details. The Multi-Scale Retinex algorithm processes the image at multiple scales and then performs weighted synthesis of the results at different scales to effectively balance the brightness and contrast of the image. bf (x, y), different Gaussian blur operations are applied at multiple scales to generate multiple blurred images. The blur degree at each scale is determined by the Gaussian convolution kernel. The standard deviation σ n Control, for each scale, calculate the Retinex output: Retinex outputs of all scales are adjusted according to the preset weight ω n Take the weighted average and get I msr (x, y); the number of scales L determines the level of detail of image processing, and 3 to 5 scales are taken to ensure that sufficient image details are captured; the weight coefficient of the scale ω n Determines the contribution of each scale in multi-scale fusion. Each weight is a positive number, and the sum of all weights is equal to 1.

[0049] The present invention is further configured to perform post-processing and color balance adjustment on the processed corrected image, and the calculation logic is: Among them, I final (x, y) is the pixel value of the final optimized image at the pixel position (x, y), λ c Add weight to the colors, is the color gradient, which is used to measure the degree of color change in an image. is the image gradient, which is used to measure the degree of change of the image pixel value; φ is the color balance control parameter, which is used to adjust the intensity of color balance; the above calculation logic is the process of post-processing and color balance adjustment of the image enhanced by the multi-scale Retinex algorithm, and the naturalness and visual consistency of the image color are ensured by color enhancement and balance adjustment of the enhanced image; the color enhancement weight λ c Adjust according to the need of image color enhancement. Higher values ​​are suitable for scenes that need to significantly enhance colors, and lower values ​​are used to maintain the natural colors of the image. The value range is [0.1, 1.0]; Color gradient It is calculated based on the difference in color channels in the image and is used to identify the degree of color change in the image. The color balance control parameter φ controls the balance between color enhancement and image gradient. A larger value makes the enhancement softer, and a smaller value makes the enhancement more significant. The value range is [0.5, 2.0]. Through color enhancement and balance adjustment, the processed image is visually more natural, the color restoration is more realistic, and the color cast problem caused by excessive enhancement is avoided.

[0050] The above embodiments can be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that contains one or more available media sets. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state hard disk.

[0051] It should be understood that the term "and / or" in this article is only a description of the association relationship of associated objects, indicating that there can be three relationships. For example, A and / or B can represent: A exists alone, A and B exist at the same time, and B exists alone. A and B can be singular or plural. In addition, the character " / " in this article generally indicates that the associated objects before and after are in an "or" relationship, but it may also indicate an "and / or" relationship. Please refer to the context for specific understanding.

[0052] In this application, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refers to any combination of these items, including any combination of single or plural items. For example, at least one of a, b, or c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0053] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0054] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0055] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0056] In the several embodiments provided in the present application, it should be understood that the disclosed system can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0057] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0058] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0059] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage media include: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks or optical disks.

[0060] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A machine vision unit used in smart community intelligent security management system, It is characterized in that include: Image acquisition module: obtains the real-time video stream of all security monitoring at night in the smart community intelligent security management system, performs frame capture processing on the real-time video stream and classifies it by scene to obtain multiple frames of images; Multi-frame image fusion module: Through the multi-exposure fusion algorithm, multiple frames of images in the same scene are fused to obtain a fused image; Adaptive correction module: Adaptive gamma correction algorithm and multi-channel image processing method are used to adaptively adjust the brightness and contrast of the fused image according to the local features of the fused image, and multi-channel detail enhancement and color correction are performed on the fused image after brightness and contrast adjustment to obtain a corrected image; Image processing module: The corrected image is processed using a bilateral filtering algorithm and a multi-scale Retinex algorithm, wherein the bilateral filtering algorithm is used to remove high-frequency noise in the corrected image, and the multi-scale Retinex algorithm is used to perform nonlinear adjustment on the brightness of the corrected image to enhance the detail performance of the corrected image; Image post-processing module: post-process and adjust the color balance of the processed corrected image to ensure the natural color and visual consistency of the image.

2. According to claim 1, a machine vision unit applied to a smart community intelligent security management system, It is characterized in that The multi-frame image fusion module includes a feature weight calculation component and a multi-exposure fusion component; The feature weight calculation component is used to preprocess multiple frames of images taken by the camera at different exposure times in the same scene, and calculate the feature weight of each image; The multi-exposure fusion component is used to fuse multiple frames of images through a multi-exposure fusion algorithm according to the feature weights.

3. According to claim 2, a machine vision unit applied to a smart community intelligent security management system, It is characterized in that The calculation logic of the feature weight of each image is: Among them, W k (x, y) is the feature weight of the k-th frame image at the pixel position (x, y), I k (x, y) is the pixel value of the k-th frame image at the pixel position (x, y), M is the median brightness of the image, σ is the brightness expansion parameter, is the image gradient, which is used to measure the degree of change of image pixel values, and α is the gradient enhancement coefficient; According to the feature weights, multiple frames of images are fused through a multi-exposure fusion algorithm, and the fusion logic is: Among them, I fusion (x, y) is the pixel value of the fused image at the pixel position x, y), N is the number of multi-frame images, and β is the nonlinear enhancement coefficient, which is used to control the nonlinear effect of fusion.

4. According to claim 1, a machine vision unit applied to a smart community intelligent security management system, It is characterized in that The adaptive correction module includes a gamma correction component and a multi-channel correction component; the gamma correction component is used to adaptively adjust the brightness and contrast of the fused image according to the local features of the fused image by using an adaptive gamma correction algorithm; the multi-channel correction component is used to perform multi-channel detail enhancement and color correction on the fused image after brightness and contrast adjustment by using a multi-channel image processing method to obtain a corrected image.

5. According to claim 4, a machine vision unit applied to a smart community intelligent security management system, It is characterized in that The calculation logic of the adaptive gamma correction algorithm is: adjust (x,y)=I fusion (x,y) γ(x,y) , where I adjust (x, y) is the pixel value of the image at the pixel position (x, y) after brightness and contrast adjustment. fusion (x, y) is the pixel value of the fused image at the pixel position (x, y), γ(x, y) is the adaptive gamma coefficient, and the calculation logic of the adaptive gamma coefficient γ(x, y) is: Among them, η is the contrast enhancement strength parameter, κ is the brightness adjustment sensitivity parameter, and T is the local brightness threshold, which is used to control the benchmark of gamma adjustment.

6. According to claim 5, a machine vision unit applied to a smart community intelligent security management system, It is characterized in that Multi-channel detail enhancement and color correction are performed on the fused image after brightness and contrast adjustment. The calculation logic is: Among them, I checkl (x, y) is the pixel value of the corrected image at the pixel position (x, y), I adjust,c (x, y) is the pixel value of the image of the cth channel of the adjusted multi-channel image at the pixel position (x, y), is the image gradient, which is used to measure the degree of change of the image pixel value, τ is a nonlinear parameter that controls the degree of detail enhancement, and c∈{R,G,B} indicates that the R, G and B color channels of the image are calculated separately.

7. According to claim 1, a machine vision unit applied to a smart community intelligent security management system, It is characterized in that The image processing module includes a bilateral filtering denoising component and a multi-scale Retinex algorithm enhancement component; the bilateral filtering denoising component is used to perform bilateral filtering on the corrected image, retain the edge features in the corrected image, and remove high-frequency noise in the corrected image; the multi-scale Retinex algorithm enhancement component is used to perform nonlinear adjustment on the brightness of the corrected image and enhance the detail performance of the corrected image.

8. According to claim 7, a machine vision unit applied to a smart community intelligent security management system, It is characterized in that The calculation logic of bilateral filtering of the corrected image is: Among them, I bf (x, y is the pixel value of the corrected image at the pixel position (x, y) after bilateral filtering, Ω is the local window range, ∥I checkl (u,v)-I checkl (x,y)∥ is the difference between the pixel values ​​at the pixel position (u,v) and the pixel value at the pixel position (x,y), ∥(u,v)-(x,y)∥ is the spatial distance between the pixel position (u,v) and the pixel position (x,y), σ r and σ s are the control parameters of pixel value difference and spatial distance respectively, and FC(x,y) is the normalization factor.

9. A machine vision unit applied to a smart community intelligent security management system according to claim 8, It is characterized in that The calculation logic for nonlinear adjustment of the brightness of the corrected image is: Among them, I msr (x, y) is the pixel value of the corrected image at the pixel position (x, y) after multi-scale Retinex algorithm enhancement, L is the number of scales, ω n is the weight coefficient of the nth scale, The scale is σ n Gaussian convolution kernel, * represents convolution operation.

10. The machine vision unit applied to the intelligent security management system of a smart community according to claim 1, It is characterized in that Post-process and adjust the color balance of the processed corrected image. The calculation logic is: Among them, I final (x, y) is the pixel value of the final optimized image at the pixel position (x, y), λ c Add weight to the colors, is the color gradient, which is used to measure the degree of color change in an image. is the image gradient, which is used to measure the degree of change of the image pixel value, and φ is the color balance control parameter, which is used to adjust the intensity of color balance.

Images