Human visual perception simulation-based self-adaptive low-illumination image enhancement method

Abstract

The invention provides a human visual perception simulation-based self-adaptive low-illumination image enhancement method. The method is put forwards based on characteristics of low brightness and low contrast of a low-illumination color image and through researching an automatic adjustment process of pupils and photoreceptor cells for an environment. The method includes the following steps that: the adjustment process of the pupils for light is simulated, and the total brightness level of the image is improved; and the self-adaptive adjustment and control ability of eye vision for a low-illumination environment is simulated, and a nonlinear mapping model is designed to simulate the adjustment process of rod cells and cone cells, so that a bright and dark self-adaptation function can be obtained, and a bright and dark information fusing function is determined according to illumination distribution, and global self-adaptive adjustment is performed on brightness components; and the local contrast of the enhanced brightness image will be decreased, so that local self-adaptive contrast enforcement is performed on the image through adopting an exponential function; and finally, and color restoration is performed on the enhanced image. With the human visual perception simulation-based self-adaptive low-illumination image enhancement method of the invention adopted, the brightness, local contrast and detail information of the low-illumination color image can be effectively improved, and especially, the method has obvious effects in dark area and highlight area enhancement.

Description

Technical field: [0001] The invention relates to image processing technology, in particular to an adaptive low-illuminance image enhancement method for simulating human visual perception. technical background: [0002] In the application of image acquisition or video surveillance, low-illumination conditions such as night, backlight, and insufficient exposure are often encountered. Inconvenience. The use of low-illuminance color image enhancement technology can effectively reduce the impact of low-illuminance environments on image or video acquisition devices, thereby effectively improving the performance of video surveillance, image recognition, computer vision and other systems under low-illumination conditions, and solving many working problems. Therefore, it is very necessary to study the enhancement of low-light images. [0003] At present, there are three main categories of low-light image enhancement algorithms: spatial domain method, transform domain method and f...

AI Technical Summary

Problems solved by technology

This method needs to keep the photographic equipment stable when shooting, takes a long time to shoot, cannot be applied to the field of real-time image and video enhancement, and has poor effect on images with very low brightness

Images