Image contrast intensifying method

Abstract

This invention relates to image contrast intensifying method, comprising the following steps: first, scanning a piece of image to sequentially calculate brightness variance of two adjacent pixels; then, comparing the variance to pre-setting threshold value, if large, generating a positive force sequence, if smaller, generating a negative force sequence; third, repeatedly operating the generation and accumulating calculation of the positive and negative ones to get a net positive force sequence and a net negative one until judging termination; fourth, forming a net force sequence by net positive one minus added-weight factor net negative one, and integrating and normalizing it to get a spreading force function; fifth, linear combining the spreading force function and original brightness switching function to get a new switching function for adjusting original brightness to intensify image contrast.

Description

technical field [0001] The invention relates to an image contrast enhancement method, in particular to an image contrast enhancement method which avoids increasing image noise and improves image unnaturalness. Background technique [0002] The light source has a very important influence on image formation. When the shooting environment is too bright or too dark, the captured image will usually be a low-contrast image. In order to improve image contrast, a commonly used basic method is histogram equalization (Histogram Equalization). When the brightness histogram of an image is obtained, a brightness transfer function is calculated according to the histogram information, and the brightness of the image is adjusted using this function, in order to achieve good image contrast quality. The basic idea of ​​histogram equalization is to assign a larger range of grayscale values ​​to the grayscale values ​​with higher values ​​in the histogram. However, this method results in loss...

AI Technical Summary

Problems solved by technology

[0004] In addition, the paper published by J.Y.Kim et al. in 2001: "An Advanced ContrastEnhancement Using Partially Overlapped Sub-Block Histogram Equalization" and the paper published by T.K.Kim et al. in 1998: "Contrast Enhancement System Using Spatially Adaptive Histogram Equalization with Electronics", proposed a method for regional histogram equalization to improve contrast, however, their proposed method is still computationally quite high in complexity

[0005] Also, the paper published by S. Sakaue et al. in 1995: "Dynamic Range Expansion of Video Cameras by Adaptive Gamma Processing" proposed to use a neuron network to improve image contrast, but training data (training data) and decision rules (decision rules) ) may affect quality as well, so it is not effective in improving contrast quality

Images