Image dehazing methods, systems, storage media, and devices based on energy domain transformation

Abstract

This invention relates to an image dehazing method, system, storage medium, and device based on energy domain transformation, comprising the following steps: Step 1: Acquire the hazy image to be processed and perform preprocessing; Step 2: Calculate the energy of each intensity in the RGB three channels of the preprocessed image, calculate threshold parameters for the energy, and limit the energy; Step 3: Convert from the energy domain back to the RGB color space and output the dehazed image. Based on the basic feature statistical information of image pixel image contrast, the energy of the image is obtained using local statistical information, and the energy of the image is limited, thereby improving the relative contrast of the energy. After conversion back to the RGB space, the local contrast in the RGB space is significantly enhanced, thus realizing complex spatial domain feature processing and achieving high-performance dehazing function.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, specifically to an image dehazing method, system, storage medium, and device based on energy domain transformation. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Due to the effects of smog, images taken in foggy weather may lose detail and have low contrast, which will affect subsequent image analysis and recognition. Similarly, water contains a large number of suspended particles that scatter light, causing a reduction in the light reflected from objects. At the same time, the reflected light mixes with the light directly received by the observer, causing changes in the contrast, sharpness, and other features of the image obtained by the observer, resulting in the loss of a lot of detailed information and also presenting a foggy image.

[0004] To address the aforementioned issues, it is necessary to analyze and preprocess hazy images to highlight details and make them more suitable for computer recognition. Current technologies mainly include image enhancement, image restoration, and convolutional neural networks, which can achieve dehazing but the results are not ideal. Summary of the Invention

[0005] To address the technical problems mentioned above, this invention provides an image dehazing method, system, storage medium, and device based on energy domain transformation. Based on the basic feature statistical information of image pixel-to-image comparison, the energy of the image is obtained using local statistical information. The energy of the image is then limited, thereby improving the relative contrast of the energy. After conversion back to RGB space, the local contrast is significantly enhanced in RGB space, thus achieving complex spatial domain feature processing and realizing high-performance dehazing.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] The first aspect of the present invention provides an image dehazing method based on energy domain transformation, comprising the following steps:

[0008] Step 1: Obtain the fogged image to be processed and perform preprocessing;

[0009] Step 2: Calculate the energy of each intensity in the RGB three channels of the preprocessed image, calculate the threshold parameter for the energy, and perform energy limiting processing;

[0010] Step 3: Convert from the energy domain back to the RGB color space and output the dehazed image.

[0011] The preprocessing in step 1 is as follows:

[0012] The image to be processed is decomposed into three energy channels: RGB.

[0013] The separated image channel range is defined within a set integer range.

[0014] Step 2, energy calculation and threshold calculation, are as follows:

[0015] Step 2.1: Based on the preprocessed image, compare each color intensity l with the surrounding pixels, and iterate through each color channel. If the intensity l of each surrounding pixel I(i,j) is greater than l, generate a point matrix B(i,j) and record it as +1; otherwise, record it as -1. Obtain a set number of basic calculation comparison pages; the formula is as follows:

[0016]

[0017] Where 'c' represents a single channel in the RGB energy channels.

[0018] Step 2.2: Calculate the image energy based on the dot matrix of a single basic page; multiply each point b(i,j) with its surrounding points, sum the results, and the image energy formula is as follows:

[0019]

[0020] Where Ω represents a block with center coordinates (i,j)3x3.

[0021] Step 2.3: Obtain the exposure value of the image channel based on the image energy, using the following formula:

[0022]

[0023] Where L is the image grayscale intensity value of 256, and l is each grayscale intensity.

[0024] Step 2.4: Divide the image region into two parts according to the exposure value, use the obtained limit value to limit the energy of the image, calculate the probability distribution function of each intensity l in the interval of each region of the image, and form a new channel image after mapping;

[0025] Specifically: The image region is divided into two parts [0, α-1] and [α, L-1], where α = (L-1) × (1 - exp), and the limit value formula is as follows:

[0026]

[0027] The energy is limited according to the following formula:

[0028]

[0029] Calculate the probability distribution function for each intensity l in the intervals [0, α-1] and [α, L-1] respectively, and obtain the cumulative distribution function CDF. After mapping, a new channel image is formed:

[0030]

[0031] In step 3, images of the three RGB channels are obtained respectively, and after fusion, the dehazed image is output.

[0032] A second aspect of the present invention provides a system for implementing the above-described method, comprising:

[0033] The preprocessing module is configured to: acquire the fogged image to be processed and perform preprocessing;

[0034] The dehazing module is configured to: calculate the energy of each intensity in the RGB three channels of the preprocessed image, calculate the threshold parameter for the energy, and perform energy limiting processing.

[0035] The preprocessing module is configured to convert the image from the energy domain back to the RGB color space and output the dehazed image.

[0036] A third aspect of the present invention provides a computer-readable storage medium.

[0037] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the energy domain transformation-based image dehazing method described above.

[0038] A fourth aspect of the present invention provides a computer device.

[0039] A computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps of the image dehazing method based on energy domain transformation as described above.

[0040] Compared with existing technologies, one or more of the above technical solutions have the following beneficial effects:

[0041] Based on the basic feature statistics of image pixel image comparison, the energy of the image is obtained by using local statistical information. The energy of the image is limited, thereby improving the relative contrast of the energy. After converting back to RGB space, the local contrast is significantly enhanced in RGB space, thus realizing complex spatial domain feature processing and achieving high-performance dehazing function. Attached Figure Description

[0042] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0043] Figure 1 This is a flowchart of an image dehazing method based on energy domain transformation provided by one or more embodiments of the present invention;

[0044] Figures 2-5 These are all images showing the dehazing effect provided by one or more embodiments of the present invention. Detailed Implementation

[0045] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0046] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0048] As described in the background section, it is necessary to analyze and preprocess hazy images to highlight the details in the images and make them more suitable for computer recognition. Current existing technologies mainly include image enhancement, image restoration, and convolutional neural networks.

[0049] Image enhancement-based methods do not consider the formation process of hazy images, but instead directly highlight the details of the image, improve contrast, and make hazy images appear clearer.

[0050] Image restoration-based methods track the physical process of image degradation and restore a clear image through a physical model.

[0051] CNN (Convolutional Neural Network) based methods utilize the powerful learning capabilities of neural networks to find the mapping relationship between fogged images and certain coefficients in the physical model of image restoration, or to use GANs to restore clear, fog-free images from fogged images.

[0052] While the above methods can achieve defogging, the results are not ideal.

[0053] Therefore, the following embodiments provide an image dehazing method, system, storage medium, and device based on energy domain transformation. Based on the basic feature statistical information of image pixel image comparison, the energy of the image is obtained using local statistical information, and the energy of the image is limited to improve the relative contrast of the energy. After conversion back to RGB space, the local contrast in RGB space is significantly enhanced, thereby realizing complex spatial domain feature processing and achieving high-performance dehazing function.

[0054] Example 1:

[0055] like Figure 1 As shown, the image dehazing method based on energy domain transformation includes the following steps:

[0056] Step 1: Obtain the fogged image to be processed and perform preprocessing;

[0057] Step 2: Calculate the energy of each intensity in the RGB three channels of the preprocessed image, calculate the threshold parameter for the energy, and perform energy limiting processing;

[0058] Step 3: Convert from the energy domain back to the RGB color space and output the dehazed image.

[0059] Specifically:

[0060] In step 1, the specific preprocessing steps are as follows:

[0061] Step 1.1: Decompose the image to be processed into three energy channels: RGB;

[0062] Step 1.2: The separated image channel intervals are defined as integer ranges within [0, 255].

[0063] In step 2, the specific steps for energy calculation and threshold calculation are as follows:

[0064] Step 2.1: Image Channel Energy Calculation: Take the color intensity l for each color channel and iteratively process each color channel. Compare l with the surrounding pixels. If the value of each surrounding pixel I(i,j) is greater than l, generate a point matrix B(i,j) and record it as +1; otherwise, record it as -1. The formula is as follows:

[0065]

[0066] Where 'c' represents a single RGB channel, and thus from 0 to 255, L=256 basic calculation comparison pages are formed.

[0067] Step 2.2: Calculate the image energy based on a single Basic Page. The image energy E is calculated from the point matrix of the Basic Page. Each point b(i,j) is multiplied by its surrounding points, and the results are summed. The image energy formula is as follows:

[0068]

[0069] Where Ω represents a block with center coordinates (i,j)3x3.

[0070] Step 2.3: Calculate the exposure value of the image channels, using the following formula:

[0071]

[0072] Where L is the image grayscale intensity value of 256, and l is each grayscale intensity.

[0073] Step 2.4: The image region is divided into two parts [0, α-1] and [α, L-1], where α = (L-1) × (1-exp). The formula for calculating the limit value is as follows:

[0074]

[0075] Then, limit the energy according to the following formula:

[0076]

[0077] Calculate the probability distribution function for each intensity l in the intervals [0, α-1] and [α, L-1] respectively, and obtain the cumulative distribution function CDF. After mapping, a new channel image is formed:

[0078]

[0079] Step 3: The RGB three channels of the image are fused [R,G,B] to form a new image.

[0080] This embodiment uses images obtained in an underwater environment as an example. Figures 2-5 The illustration shows the effect diagram of the original image and the dehazed image after processing according to this embodiment.

[0081] The above method is based on the basic feature statistics of image pixel image comparison. It uses local statistical information to obtain the image energy, limits the image energy, thereby improving the relative contrast of energy. After converting back to RGB space, it significantly enhances the local contrast in RGB space, thus realizing complex spatial domain feature processing and achieving high-performance dehazing function.

[0082] Example 2:

[0083] A system for implementing the above method includes:

[0084] The preprocessing module is configured to: acquire the fogged image to be processed and perform preprocessing;

[0085] The dehazing module is configured to: calculate the energy of each intensity in the RGB three channels of the preprocessed image, calculate the threshold parameter for the energy, and perform energy limiting processing.

[0086] The preprocessing module is configured to convert the image from the energy domain back to the RGB color space and output the dehazed image.

[0087] The system is based on the basic feature statistics of image pixel image comparison. It uses local statistical information to obtain the image energy, limits the image energy, thereby improving the relative contrast of energy. After converting back to RGB space, it significantly enhances the local contrast in RGB space, thus realizing complex spatial domain feature processing and achieving high-performance dehazing function.

[0088] Example 3:

[0089] This embodiment provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps in the image dehazing method based on energy domain transformation as described in Embodiment 1 above.

[0090] The image dehazing method system based on energy domain transformation obtains the image energy using local statistical information through basic feature statistics of image pixel-to-image comparison, limits the image energy, thereby improving the relative energy contrast, and significantly enhances the local contrast in RGB space after conversion back to RGB space. This achieves complex spatial domain feature processing and realizes high-performance dehazing function.

[0091] Example 4:

[0092] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the steps in the image dehazing method based on energy domain transformation as described in Embodiment 1 above.

[0093] The image dehazing method system based on energy domain transformation obtains the image energy using local statistical information through basic feature statistics of image pixel-to-image comparison, limits the image energy, thereby improving the relative energy contrast, and significantly enhances the local contrast in RGB space after conversion back to RGB space. This achieves complex spatial domain feature processing and realizes high-performance dehazing function.

[0094] The steps or modules involved in Embodiments 2 to 4 above correspond to those in Embodiment 1. For specific implementation details, please refer to the relevant description section of Embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood as including any medium capable of storing, encoding, or carrying an instruction set for execution by a processor and enabling the processor to perform any of the methods in this invention.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An image defogging method based on energy domain transformation, characterized in that: The method comprises the following steps: Step 1: obtaining a foggy image to be processed and performing preprocessing; Step 2: performing energy calculation of each intensity on the RGB three channels of the preprocessed image respectively, performing threshold parameter calculation on the energy, and performing limiting processing on the energy; Step 2.1: Based on the pre-processed image, take each gray intensity l , and iterate each color channel, compare with the surrounding pixels, if each pixel I( i , j ) is greater than l , generate dot b( i , j ) as +1, otherwise as -1, get a set number of basic calculation contrast basic page; formula as follows: Wherein c represents a single channel in the RGB energy channel; Step 2.2: Calculate the energy of the image based on the dot product of the single basic page; specifically, each point b( i , j ) is multiplied by the surrounding points, and the sum is accumulated to obtain the energy of the image; the image energy formula is as follows: where Ω represents a 3x3 block with center coordinates (x, y) and i , j )3x3 block. Step 2.3: obtaining an exposure value of the image channel based on the energy of the image; the formula is as follows: where L is the value 256 of the image gray intensity, l for each gray intensity; Step 2.4: According to the exposure value, the image region is divided into two parts, the energy of the image is limited by the obtained limiting value, and the probability distribution function of each gray intensity in the interval of each region of the image is calculated respectively to form a new channel image after mapping; Specifically, the region of the image is divided into two parts [0, a-1] and [a, L-1], where a=(L-1)×(1-exp), and the limiting value formula is as follows: l Step 2.4: According to the exposure value, the image region is divided into two parts, the energy of the image is limited by the obtained limiting value, and the probability distribution function of each gray intensity in the interval of each region of the image is calculated respectively to form a new channel image after mapping; Specifically, the region of the image is divided into two parts [0, a-1] and [a, L-1], where a=(L-1)×(1-exp), and the limiting value formula is as follows: where L is the value 256 of the image gray intensity, l for each gray intensity; The energy is limited according to the following formula: Calculate the gray intensity for each value in the intervals [0, α-1] and [α, L-1]. l The probability distribution function is obtained, thus the cumulative distribution function (CDF) is derived, and after mapping, a new channel image is formed: Step 3: converting back from the energy domain to the RGB chrominance space and outputting the image after fog removal. 2.The energy domain transform based image defogging method of claim 1, wherein: The step 1 is specifically: Decomposing the image to be processed into RGB three energy channels; The interval of the separated image channel is defined as an integer range.

3. The image defogging system based on energy domain transform, adopting the image defogging method based on energy domain transform according to any one of claims 1-2, characterized in that: It comprises: A preprocessing module configured to obtain a foggy image to be processed and perform preprocessing; A fog removal module configured to perform energy calculation of each intensity on the RGB three channels of the preprocessed image respectively, perform threshold parameter calculation on the energy, and perform limiting processing on the energy; The preprocessing module is configured to convert back from the energy domain to the RGB chrominance space and output the image after fog removal.

4. A computer readable storage medium having stored thereon a computer program, which, when executed by a processor, implements the steps in the image fog removal method based on energy domain transformation according to any one of claims 1-2.

5. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps in the image fog removal method based on energy domain transformation according to any one of claims 1-2 when executing the program.

Images