CN-116342430-B - Enhancement method, device and equipment for low-brightness image

CN116342430BCN 116342430 BCN116342430 BCN 116342430BCN-116342430-B

Abstract

The invention discloses a method, a device and equipment for enhancing a low-brightness image, the method comprises the steps of inputting the image to be processed into a pre-built image decomposition model, restraining illumination gradient in the image to be processed by using the image decomposition model to obtain a first illumination component, performing fidelity on reflection gradient in the image to be processed by using the image decomposition model to obtain a first reflection component, adjusting illumination intensity of the first illumination component according to a pre-determined adjusting rule to generate a second illumination component, and generating an enhanced image based on the second illumination component and the first reflection component.

Inventors

GAO YIN
LI JUN
YAN CHAO

Assignees

泉州装备制造研究所

Dates

Publication Date: 20260508
Application Date: 20230406

Claims (9)

1. A method of enhancing a low-luminance image, comprising: Inputting an image to be processed into a pre-constructed image decomposition model, utilizing the image decomposition model to restrict illumination gradients in the image to be processed to obtain a first illumination component, and utilizing the image decomposition model to perform fidelity on reflection gradients in the image to be processed to obtain a first reflection component; adjusting the illumination intensity of the first illumination component according to a preset adjustment rule to generate a second illumination component; generating an enhanced image based on the second illumination component and the first reflection component; the adjusting the illumination intensity of the first illumination component according to a predetermined adjustment rule to generate a second illumination component includes: determining an adjustment parameter corresponding to each pixel in the image to be processed based on the adjustment rule, wherein the adjustment rule is a mapping relation between the adjustment parameter and the pixel; adjusting the illumination intensity of the first illumination component based on all the adjustment parameters to generate second illumination intensity; the regulation rule is embodied in the form of a regulation function, and the regulation function is an exponential function: wherein, the method comprises the steps of, For the adjustment function, x is different pixel values, and a, b and d are fixed parameters; the enhanced image is: wherein, the method comprises the steps of, For the enhanced image to be used in the present invention, For the first reflected component of the light beam, For the first component of the illumination, For the second component of the illumination, For the purpose of the regulation rules, Is an initial illumination component; the initial illumination component is determined based on gray values of the image to be processed at R, G, B three channels.
2. The method of enhancing a low-luminance image according to claim 1, wherein the image decomposition model comprises: a first fidelity sub-model, a second fidelity sub-model, and a constraint sub-model.
3. The method of enhancing a low-luminance image according to claim 2, wherein obtaining the first illumination component and the first reflection component comprises: Constructing a first image decomposition sub-model based on the first fidelity sub-model and the constraint sub-model, and constructing a second image decomposition sub-model based on the first fidelity sub-model and the second fidelity sub-model; Inputting the initial illumination component into the first image decomposition sub-model to obtain a third illumination component; inputting the third illumination component into the second image decomposition sub-model to obtain a second reflection component; And when the third illumination component meets a first preset condition and the second reflection component meets a second preset condition, determining the third illumination component as the first illumination component and determining the second reflection component as the first reflection component.
4. A method of enhancing a low intensity image as claimed in claim 3 wherein the second image decomposition sub-model constrains the reflection gradient in the image to be processed with a preset enhancement image gradient.
5. A method of enhancing a low intensity image as claimed in claim 3, wherein said constructing a first image decomposition model based on said first fidelity sub-model and said constraint sub-model comprises: Constructing a third image decomposition sub-model based on the first fidelity sub-model and the constraint sub-model; Simplifying a first fidelity sub-model in the third image decomposition sub-model into a third fidelity sub-model formed from the initial illumination component; And replacing the constraint sub-model in the third image decomposition sub-model with a third fidelity sub-model equivalent to the constraint sub-model to generate a first image decomposition sub-model.
6. The method of enhancing a low-luminance image according to any one of claims 3 to 5, wherein the first image decomposition sub-model includes a first decomposition term and a second decomposition term, and wherein the inputting the initial illumination component into the first image decomposition sub-model results in a third illumination component, comprising: Inputting the initial illumination component and a predefined constraint condition into the first decomposition term to obtain an auxiliary variable, wherein the auxiliary variable is used for constraining the illumination gradient; and inputting the auxiliary variable, the initial illumination component and a first preset coefficient into the second analysis item to obtain the third illumination component, wherein the first preset coefficient is used for representing the similarity between the constraint condition and the auxiliary variable.
7. The method of enhancing a low-luminance image according to claim 6, further comprising: Updating the iteration times when the third illumination component does not meet the first preset condition and/or the second reflection component does not meet the second preset condition; comparing the updated iteration times with a preset iteration threshold; When the updated iteration number reaches the preset iteration threshold, determining the third illumination component as the first illumination component, and determining the second reflection component as the first reflection component; Or when the updated iteration times do not reach the preset iteration threshold, adjusting the first preset coefficient to a second preset coefficient according to a preset rule; inputting the auxiliary variable, the initial illumination component and the second preset coefficient into the second decomposition term, and inputting the result output by the second decomposition term into the second image decomposition sub-model; Stopping the iterative operation until the finally output illumination component meets the first preset condition and the output reflection component meets the second preset condition or when the iteration number reaches the preset iteration threshold, determining the finally output illumination component as the first illumination component, and determining the finally output reflection component as the first reflection component.
8. An enhancement device for a low-luminance image, comprising: The input module is used for inputting an image to be processed into a pre-constructed image decomposition model, restraining the illumination gradient in the image to be processed by using the image decomposition model to obtain a first illumination component, and performing fidelity on the reflection gradient in the image to be processed by using the image decomposition model to obtain a first reflection component; the adjusting module is used for adjusting the illumination intensity of the first illumination component according to a preset adjusting rule to generate a second illumination component; A generation module for generating an enhanced image based on the second illumination component and the first reflection component; the adjusting the illumination intensity of the first illumination component according to a predetermined adjustment rule to generate a second illumination component includes: determining an adjustment parameter corresponding to each pixel in the image to be processed based on the adjustment rule, wherein the adjustment rule is a mapping relation between the adjustment parameter and the pixel; adjusting the illumination intensity of the first illumination component based on all the adjustment parameters to generate second illumination intensity; the regulation rule is embodied in the form of a regulation function, and the regulation function is an exponential function: wherein, the method comprises the steps of, For the adjustment function, x is different pixel values, and a, b and d are fixed parameters; the enhanced image is: wherein, the method comprises the steps of, For the enhanced image to be used in the present invention, For the first reflected component of the light beam, For the first component of the illumination, For the second component of the illumination, For the purpose of the regulation rules, Is an initial illumination component; the initial illumination component is determined based on gray values of the image to be processed at R, G, B three channels.
9. A computer device comprising a memory and a processor, the memory and the processor being communicatively coupled to each other, the memory being configured to store a computer program which, when executed by the processor, causes the processor to perform the method of enhancing a low intensity image as claimed in any one of claims 1 to 7.

Description

Enhancement method, device and equipment for low-brightness image Technical Field The present invention relates to the field of image processing, and in particular, to a method, an apparatus, and a device for enhancing a low-brightness image. Background In a real scene, the problems of light, visual angle and the like can lead to that pictures shot by us are relatively dark, and the dark pictures not only can influence the observation of us, but also can greatly influence the effect of a computer vision processing algorithm. Based on this, a low-luminance image enhancement method occurs. In the related art, an image brightness enhancement method generally decomposes an input image into an illumination component and a reflection component based on a Retinex model, and then corrects global enhancement illumination using Gamma, thereby achieving brightness enhancement of the image. However, the constraint condition adopted by the current enhancing method based on the Retinex model is insufficient to obtain accurate illumination component and reflection component, so that the problems of color distortion, halation artifact and the like appear in the finally enhanced image. Disclosure of Invention Therefore, the technical problem to be solved by the invention is to overcome the defect that the illumination component and the reflection component cannot be accurately decomposed in the prior art, so as to provide a method, a device and equipment for enhancing the low-brightness image. In a first aspect, the present invention provides a method for enhancing a low-luminance image, comprising: Inputting an image to be processed into a pre-constructed image decomposition model, restraining an illumination gradient in the image to be processed by using the image decomposition model to obtain a first illumination component, performing fidelity on a reflection gradient in the image to be processed by using the image decomposition model to obtain a first reflection component, adjusting the illumination intensity of the first illumination component according to a pre-determined adjustment rule to generate a second illumination component, and generating an enhanced image based on the second illumination component and the first reflection component. The invention utilizes the pre-constructed image decomposition model to constrain the illumination gradient and perform fidelity on the reflection gradient, so that the finally obtained first illumination component and first reflection component corresponding to the image to be processed are more accurate. And adjusting the illumination intensity of the first illumination component through a predetermined adjustment rule to obtain a second illumination component. Then, an enhanced image is generated based on the second illumination component and the first reflection component. The enhanced image obtained through the mode can keep the visibility of the local high-intensity area while enhancing the weak light area, so that the robot can have good visibility in a complex weak light environment, and the working efficiency of equipment can be greatly improved. With reference to the first aspect, in a first embodiment of the first aspect, the image decomposition model includes: a first fidelity sub-model, a second fidelity sub-model, and a constraint sub-model. With reference to the first aspect, in a second embodiment of the first aspect, obtaining the first illumination component and the first reflection component includes: The method comprises the steps of determining initial illumination components based on gray values of three channels of an image to be processed in R, G, B, constructing a first image decomposition sub-model based on the first fidelity sub-model and the constraint sub-model, constructing a second image decomposition sub-model based on the first fidelity sub-model and the second fidelity sub-model, inputting the initial illumination components into the first image decomposition sub-model to obtain third illumination components, inputting the third illumination components into the second image decomposition sub-model to obtain second reflection components, and determining the third illumination components as the first illumination components and the second reflection components as the first reflection components when the third illumination components meet a first preset condition and the second reflection components meet a second preset condition. With reference to the first aspect, in a third embodiment of the first aspect, the second image decomposition sub-model uses a preset enhanced image gradient to constrain a reflection gradient in the image to be processed. With reference to the first aspect, in a fourth embodiment of the first aspect, the constructing a first image decomposition model based on the first fidelity sub-model and the constraint sub-model includes: The method comprises the steps of constructing a first image decomposition sub-model based on the