Rainy day image processing system based on intelligent image algorithm

Abstract

The application relates to the technical field of image processing, and discloses a rain day image processing system based on an intelligent image algorithm, which comprises a vanguard data acquisition module, a vanguard data processing module, a result storage module, a rain day image acquisition module, a humidity information acquisition module, a core processing module and a real-time feedback module; the vanguard data acquisition module is used for acquiring normal image data in a recent period of time, selecting image data meeting a condition from the normal image data, and acquiring humidity data corresponding to the image data meeting the condition; the vanguard data is analyzed and processed to obtain a brightness threshold value and a humidity threshold value; when rain day image processing is carried out, raindrops in the image can be accurately recognized by using a three-dimensional coordinate system and an RGB value; the raindrops in the image are removed by using a pixel point filling method; and water vapor in the image is removed by using the humidity threshold value, so that the whole image becomes clearer.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and more specifically discloses a rainy day image processing system based on intelligent image algorithms. Background Technology

[0002] With the development of modern technology, cameras have become an important tool for public safety protection, capturing video images and sending them to relevant authorities to ensure public safety. However, during rainy weather, increased moisture in the air and a large number of raindrops can blur the images captured by cameras, affecting viewers' ability to see the pictures. Current technologies typically use neural network algorithms to process these images. However, this method requires training the neural network model, which necessitates a large dataset. This results in a long system deployment cycle. Furthermore, when processing rainy images, neural network algorithms can only remove raindrops; they cannot improve the image clarity, leading to low-quality processed images. Summary of the Invention

[0003] The main technical problem solved by this invention is to provide a rainy day image processing system based on intelligent image algorithms, which can solve the problems of current rainy day image processing methods, such as excessively long investment cycles and low image quality after processing.

[0004] To solve the above-mentioned technical problems, according to one aspect of the present invention, more specifically a rainy day image processing system based on intelligent image algorithms, the system includes: a preliminary data acquisition module, a preliminary data processing module, a result storage module, a rainy day image acquisition module, a humidity information acquisition module, a core processing module, and a real-time feedback module.

[0005] Preliminary data acquisition module: used to acquire normal image data within a recent period, select image data that meets the criteria, and acquire humidity data corresponding to the image data that meets the criteria.

[0006] Preliminary data processing module: Used to process and analyze the data collected by the preliminary data acquisition module and obtain the results;

[0007] Results storage module: Used to store the analysis and processing results obtained by the preliminary data processing module;

[0008] Rainy Day Image Acquisition Module: Used to acquire and collect images of rainy days;

[0009] Humidity information acquisition module: used to acquire humidity data corresponding to the current rainy day image;

[0010] Core processing module: Used to process rainy day images acquired by the rainy day acquisition module based on the data stored in the structure storage module and the data acquired by the humidity information acquisition module;

[0011] Real-time feedback module: Used to provide real-time feedback of rainy day image data processed by the core processing module to the viewer.

[0012] Furthermore, the preliminary data acquisition module includes: a time information acquisition module, a normal image acquisition module, an image selection module, and a corresponding humidity acquisition module;

[0013] Time information acquisition module: used to acquire information about the most recent period of time.

[0014] Normal image acquisition module: used to acquire several normal image data based on the information obtained by the time information acquisition module within a recent period;

[0015] Image selection module: used to select multiple image data that meet certain conditions from several normal image data acquired by the normal image acquisition module;

[0016] Corresponding humidity acquisition module: used to acquire humidity data corresponding to multiple images after the image selection module selects multiple images that meet the conditions.

[0017] Furthermore, the preliminary data processing module includes: a brightness threshold acquisition module and a humidity threshold acquisition module;

[0018] Brightness threshold acquisition module: used to automatically acquire the brightness threshold data of the image based on multiple image data selected by the image selection module;

[0019] Humidity threshold acquisition module: This module is used to automatically obtain the humidity threshold data of the image based on multiple humidity data obtained by the corresponding humidity acquisition module.

[0020] Furthermore, the specific process by which the brightness threshold acquisition module obtains the brightness threshold data is as follows: First, it acquires the brightness information data of each pixel in each image, and then calculates the brightness threshold data of each image based on the brightness information data of each pixel in each image.

[0021]

[0022] Among them, Bh t Z represents the brightness threshold data for each image. px Bh represents the total number of pixels in each image. i The brightness data for each pixel i;

[0023] Then, the overall brightness threshold data of the image data is calculated based on the brightness threshold data of each image data:

[0024]

[0025] in, Z represents the overall brightness threshold data for the image data. img The total number of images, (Bh) t ) j The brightness threshold data for each image data j.

[0026] Furthermore, the specific process by which the humidity threshold acquisition module obtains humidity threshold data is as follows: First, the image selection module selects several images that meet the conditions, and then calculates the overall humidity threshold data of the image data based on the acquired multiple humidity information data.

[0027]

[0028] Among them, Hd t The overall humidity threshold data for the image data, Hd n The humidity information data corresponding to each image data j.

[0029] Furthermore, the core processing module includes: a three-dimensional coordinate system establishment module, a judgment and recognition module, a removal module, and an enhancement module;

[0030] 3D coordinate system establishment module: used to establish a 3D coordinate system with any corner point of the rainy day image obtained by the rainy day image acquisition module as the origin;

[0031] Judgment and Recognition Module: Used to judge and recognize raindrops in the image based on the established three-dimensional coordinate system and RGB information data;

[0032] Removal module: Based on the judgment and recognition results from the judgment and recognition module, it removes raindrops from the image and also removes water vapor from the image.

[0033] Enhancement module: Used to enhance the image after the raindrops have been removed by the removal module.

[0034] Furthermore, the removal process of the removal module is as follows: First, the pixel of each raindrop is acquired, and the RGB data of the adjacent pixels of each raindrop are used to fill the pixel of the current raindrop. At the same time, the humidity information data collected by the humidity information acquisition module and the humidity threshold data stored in the result storage module are acquired, and the removal of raindrops and water vapor in the rainy day image is completed through the following conditional expression:

[0035]

[0036] in, Each represents the current raindrop pixel (px). i RGB values, Each is the current pixel (px). i Adjacent pixels (px) i+1 RGB values, Hd n Hd represents the humidity information corresponding to the rainy day image. t This is humidity threshold data.

[0037] Furthermore, the enhancement module's enhancement process for rainy day images involves: acquiring the brightness value data of each pixel in the image after the removal module has removed raindrops and water vapor. Simultaneously, brightness threshold data is retrieved from the results storage module. Then, the enhancement value F is obtained based on the acquired data:

[0038]

[0039] The obtained enhancement values ​​are then used to enhance each pixel (px) of the rainy day image. i Enhancement processing can be performed to ultimately enhance the entire image;

[0040] in, For each pixel in the image (px) i Brightness data, This is the brightness threshold data. The maximum brightness value among the pixels in the image. The minimum brightness value among the pixels in the image.

[0041] Furthermore, the real-time feedback module includes: a wireless transmission module and an image data storage module;

[0042] Wireless transmission module: Used to send the processed rainy day image data to viewers in real time for viewing;

[0043] Image data storage module: Used to store the processed rainy day image data.

[0044] The beneficial effects of the rainy day image processing system based on intelligent image algorithms of this invention are as follows: This invention obtains brightness and humidity thresholds by analyzing and processing preliminary data. Then, during rainy day image processing, it accurately identifies raindrops in the image by establishing a three-dimensional coordinate system and using RGB values. Raindrops are then removed by pixel filling, and moisture is removed using the humidity threshold, making the entire image clearer. Finally, by combining the brightness threshold with the brightness data of each pixel after removing raindrops and moisture, the image is enhanced, making it even clearer and brighter, thus improving viewer experience. Compared to existing rainy day image processing methods, this invention processes each pixel of the rainy day image from a microscopic pixel perspective, treating the surface from a point, resulting in a clearer and higher-quality image. Furthermore, during the initial preparation process, the system automatically acquires several images from the recent period and selects multiple undamaged images from them. At the same time, it acquires the corresponding humidity data, thereby calculating the brightness threshold data and humidity threshold data. This eliminates the need for a large amount of data to support training, shortens the overall system deployment cycle, and facilitates the rapid deployment of the system. Attached Figure Description

[0045] The present invention will now be described in further detail with reference to the accompanying drawings and specific implementation methods.

[0046] Figure 1 This is a schematic diagram of the system principle. Detailed Implementation

[0047] The present invention will be described in detail below with reference to the accompanying drawings and embodiments. It should be noted that, unless otherwise specified, the embodiments and features described in the present application can be combined with each other.

[0048] According to one aspect of the invention, such as Figure 1 As shown, a rainy day image processing system based on intelligent image algorithms is provided, including:

[0049] The preliminary data acquisition module is used to acquire normal image data from a recent period, select images that meet certain criteria, and simultaneously acquire the corresponding humidity data for these images. This module includes: a time information acquisition module (for acquiring information from the most recent period); a normal image acquisition module (for acquiring several normal image data points based on the recent information acquired by the time information acquisition module); an image selection module (for selecting multiple images that meet certain criteria from the normal image data acquired by the normal image acquisition module); and a corresponding humidity acquisition module (for acquiring humidity data corresponding to the multiple images selected by the image selection module). For example, it acquires all images from the past two months that are free from damage or mosaic effects, then selects images that meet criteria such as high visibility and clarity, and simultaneously acquires the corresponding humidity information data (obtained via a humidity sensor) for these images.

[0050] The preliminary data processing module processes and analyzes the data acquired by the preliminary data acquisition module and obtains the results. This module includes: a brightness threshold acquisition module, which automatically acquires the brightness threshold data of multiple images selected by the image selection module. Specifically, it first acquires the brightness information data of each pixel in each image, and then calculates the brightness threshold data for each image based on this data.

[0051]

[0052] Among them, Bh t Z represents the brightness threshold data for each image. px Bh represents the total number of pixels in each image. i The brightness data for each pixel i;

[0053] Then, the overall brightness threshold data of the image data is calculated based on the brightness threshold data of each image data:

[0054]

[0055] in, Z represents the overall brightness threshold data for the image data. img The total number of images, (Bh) t ) j Brightness threshold data for each image data j;

[0056] Humidity threshold acquisition module: This module automatically acquires the humidity threshold data of an image based on multiple humidity data points obtained from the corresponding humidity acquisition module. The specific process is as follows: First, the image selection module selects several images that meet the specified conditions, along with their corresponding humidity information. Then, based on the acquired humidity information, the overall humidity threshold data for the image data is calculated.

[0057]

[0058] Among them, Hd t The overall humidity threshold data for the image data, Hd n The humidity information data corresponding to each image data j.

[0059] Result storage module: Used to store the brightness threshold data and humidity threshold data obtained by the preliminary data processing module.

[0060] Rainy Day Image Acquisition Module: Used to acquire and collect images of rainy days. It acquires rainy day images via a camera.

[0061] Humidity information acquisition module: Used to acquire humidity data corresponding to the current rainy day image. It acquires the current humidity information data through a sensor.

[0062] Core processing module: Used to process rainy day images acquired by the rainy day acquisition module based on the data stored in the structure storage module and the data acquired by the humidity information acquisition module. This module includes: a 3D coordinate system establishment module, used to establish a 3D coordinate system with any corner point of the rainy day image acquired by the rainy day image acquisition module as the origin; a judgment and recognition module, used to judge and recognize raindrops in the image based on the established 3D coordinate system and RGB information data; for example, in the image, a raindrop is a white line segment, which exists only on a two-dimensional plane and has no information in another dimension. Therefore, raindrops can be accurately identified by their white RGB values ​​and the fact that they exist only in two dimensions; and a removal module, used to remove raindrops from the image based on the judgment and recognition results; and to remove water vapor from the image. The removal process is as follows: first, the pixel of each raindrop is acquired, and the RGB data of the adjacent pixels of each raindrop are used to fill the pixel of the current raindrop. The filling process is to fill from the outer pixels of the raindrop inward. At the same time, the humidity information data acquired by the humidity information acquisition module and the humidity threshold data stored in the result storage module are acquired, and the removal of raindrops and water vapor from the rainy day image is completed through the following conditional expression:

[0063]

[0064] in, Each represents the current raindrop pixel (px). i RGB values, Each is the current pixel (px). i Adjacent pixels (px) i+1 RGB values, Hd n Hd represents the humidity information corresponding to the rainy day image. t This is humidity threshold data;

[0065] Enhancement module: After the raindrop removal module removes raindrops from the image, it enhances the image by acquiring the brightness value data of each pixel in the image after the raindrop and water vapor removal module has removed them. Simultaneously, brightness threshold data is retrieved from the results storage module. Then, the enhancement value F is obtained based on the acquired data:

[0066]

[0067] The obtained enhancement values ​​are then used to enhance each pixel (px) of the rainy day image. i Enhancement processing can be performed to ultimately enhance the entire image;

[0068] in, For each pixel in the image (px) i Brightness data, This is the brightness threshold data. The maximum brightness value among the pixels in the image. The minimum brightness value among the pixels in the image.

[0069] Real-time feedback module: This module is used to provide real-time feedback of rainy day image data processed by the core processing module to the viewer. This module includes: a wireless transmission module: used to send the processed rainy day image data to the viewer in real time; and an image data storage module: used to store the processed rainy day image data.

[0070] All electrical components mentioned in this article are real-world electrical components.

[0071] Of course, the above description is not a limitation of the present invention, and the present invention is not limited to the examples given above. Any changes, modifications, additions or substitutions made by those skilled in the art within the scope of the present invention are also within the protection scope of the present invention.

Claims

1. A rainy-day image processing system based on intelligent image algorithms, characterized in that, Including: An advance data acquisition module, an advance data processing module, a result storage module, a rainy-day image acquisition module, a humidity information collection module, a core processing module, and a real-time feedback module; Advance data acquisition module: Used to acquire normal image data within a recent period of time, select qualified image data therefrom, and at the same time acquire the humidity data corresponding to the qualified image data; This module includes: a time information acquisition module, a normal image acquisition module, an image selection module, and a corresponding humidity acquisition module; Advance data processing module: Used to process and analyze the data acquired by the advance data acquisition module and obtain results; This module includes: a brightness threshold acquisition module, a humidity threshold acquisition module; Brightness threshold acquisition module: Used to automatically acquire the brightness threshold data of the image according to multiple image data selected by the image selection module; The specific process for the brightness threshold acquisition module to acquire the brightness threshold data is: First, acquire the brightness information data of each pixel point in each image, and then calculate the brightness threshold data of each image according to the brightness information data of each pixel point in each image: ； Among them, is the brightness threshold data for each image, is the total number of pixel points in each image, is the brightness data of each pixel point i; Then, calculate the overall brightness threshold data of the image data according to the brightness threshold data of each image data: ； Among them, is the overall brightness threshold data of the image data, is the total amount of the images, is the brightness threshold data of each image data j; Humidity threshold acquisition module: Used to automatically acquire the humidity threshold data of the image according to multiple humidity data acquired by the corresponding humidity acquisition module; The specific process for the humidity threshold acquisition module to acquire the humidity threshold data is: First, acquire the humidity information data corresponding to several image data that meet the conditions selected by the image selection module, and then calculate the overall humidity threshold data of the image data according to the multiple humidity information data acquired: ； Among them, is the overall humidity threshold data of the image data, is the humidity information data corresponding to each image data j; Result storage module: Used to store the analysis and processing results obtained by the advance data processing module; Rainy-day image acquisition module: Used to acquire and collect rainy-day images; Humidity information collection module: Used to collect the humidity data corresponding to the current rainy-day image; Core processing module: Used to process the rainy-day images acquired by the rainy-day acquisition module according to the data stored in the structure storage module and the data acquired by the humidity information collection module; Real-time feedback module: Used to real-time feedback the rainy-day image data processed by the core processing module to the viewer.

2. The rainy-day image processing system based on an intelligent image algorithm according to claim 1, wherein: Time information acquisition module: Used to acquire the time information of the most recent period of time from now; Normal image acquisition module: Used to acquire several normal image data according to the most recent time information acquired by the time information acquisition module; Image selection module: Used to select multiple qualified image data from several normal image data acquired by the normal image acquisition module; Corresponding humidity acquisition module: Used to collect and acquire the humidity data corresponding to multiple images after the image selection module selects multiple qualified images.

3. The rainy-day image processing system based on the intelligent image algorithm according to claim 1, wherein: The core processing module includes: a three-dimensional coordinate system establishment module, a judgment and recognition module, a removal module, and an enhancement module; Three-dimensional coordinate system establishment module: used to establish a three-dimensional coordinate system with any corner point of the rainy-day image obtained by the rainy-day image acquisition module as the origin; Judgment and recognition module: used to judge and recognize raindrops in the image according to the established three-dimensional coordinate system and RGB information data; Removal module: used to remove the raindrops in the image according to the result judged and recognized by the judgment and recognition module; at the same time, remove the water vapor in the image; Enhancement module: used to perform enhancement processing on the image after removing raindrops after the raindrops in the image are removed by the removal module.

4. The rainy-day image processing system based on intelligent image algorithms according to claim 3, wherein: The removal process of the removal module is specifically as follows: first, obtain the pixel points of each raindrop, and fill the pixel points of the current raindrop with the RGB data of the adjacent pixel points of each pixel point. At the same time, obtain the humidity information data collected by the humidity information acquisition module and the humidity threshold data stored in the result storage module, and complete the removal of raindrops and water vapor in the rainy-day image through the following conditional formula: ； Among them, are the RGB values of the current raindrop pixel points respectively, are the adjacent pixel points of the current pixel point respectively, and are the RGB values of the adjacent pixel points is the humidity information corresponding to the rainy day image, is the humidity threshold data.

5. The rainy-day image processing system based on an intelligent image algorithm according to claim 3, wherein: The enhancement process of the enhancement module for rainy-day images is as follows: Obtain the brightness value data of each pixel point in the image after the removal module removes raindrops and water vapor from the image , and at the same time obtain the brightness threshold data from the result storage module , and then obtain the enhancement value F according to the obtained data: ； And use the obtained enhancement value to process each pixel of the rainy-day image for enhancement processing, and finally the entire image can be enhanced; wherein, is the luminance data of each pixel of the image , is the luminance threshold data is the maximum luminance value among the image pixels is the minimum luminance value among the image pixels.

6. The rainy-day image processing system based on an intelligent image algorithm according to claim 1, wherein: The real-time feedback module includes: a wireless transmission module, an image data storage module; Wireless transmission module: used to transmit the processed rainy-day image data to the viewer for viewing in real time; Image data storage module: used to store the processed rainy-day image data.

Images