A Low Illumination Image Enhancement Method Using Adaptive Correction

Abstract

The invention relates to a method for enhancing a low-illumination image by adopting self-adaptive correction, and belongs to the field of image processing. A single low-illumination grey level image is processed to make the visual perception of the single low-illumination grey level image clear. The method comprises the following steps of first, evaluating an illuminating component of a low-illumination gray level image V by adopting maximum filtering, minimum filtering and edge-preservation filtering in sequence, so as to obtain a reflecting component; afterwards, dividing the illuminating component into dark parts and bright parts through an otsu method, calculating the mean value of the dark parts, so as to obtain a threshold value for correction, and carrying out self-adaptive logarithmic correction on the illuminating component according to the threshold value; finally, multiplying the corrected illuminating component and the reflecting component, so as to obtain an enhanced image. The method provided by the invention can be used for promoting global brightness; the enhanced image is clear in detail and quick in processing speed, and has a quite good visual effect.

Description

technical field [0001] The invention relates to the field of image processing, in particular to a low-illuminance image enhancement method using adaptive correction, which can be used for rapid enhancement of a single image or video. Background technique [0002] With the development of modern computers, computer vision systems have been more widely used, such as video surveillance, urban traffic, satellite remote sensing, etc., especially with the widespread popularity of mobile terminals, photo collection has become within reach. However, in the case of low illumination or underexposure, it often leads to degradation problems such as low brightness, fuzzy details, low color saturation and color cast in the captured image, which brings many difficulties and challenges to the computer vision system. Therefore, it is necessary to enhance the processing and research of low-illuminance color images. [0003] As early as 1971, American scientist Land and others proposed the imp...

AI Technical Summary

Problems solved by technology

[0004] Generally, the processing of low-light images is carried out in the RGB color space. Different ratios of R (red), G (green), and B (blue) can be combined into different colors. If R, G, and B are respectively processed Processing, and then combined, will inevitably cause color distortion

Images