低信噪比图像子像素细匹配方法分析

摘要

This paper proposes a method for fine matching of image sub-pixel with low signal-to-noise ratio based on hybrid evolutionary algorithm.The sub-band with high frequency and low frequency of image is obtained,and nonlinear gain function is used to process coefficient of the sub-band with high frequency,then fine matching for edge detail of fuzzy region of image is carried out while suppressed image noise to carry out fine matching for the sub -band with low frequency.The fuzzy entropy and visual factor of image are described to obtain the optimization objective function of the fine matching,and matching parameter is selected according to the hybrid evolutionary algorithm.Thus,the fine matching with high precision is achieved.Experimental results show that the method can improve visual effect of image and has robustness.%对低信噪比图像子像素的细匹配,能够有效提高图像采集质量,增强图像视觉效果.对图像的子像素进行细匹配处理,需要获得像素匹配优化目标函数,再依据混合进化算法选取匹配参数,完成对低信噪比图像的子像素细匹配.传统方法计算图像直方图相应的模糊逻辑值,得到图像亮度分量的细匹配值,但忽略了获取图像像素匹配参数,导致匹配精度偏低.提出基于混合进化算法的低信噪比图像子像素细匹配处理方法.获得图像高频子带和低频子带,利用非线性增益函数处理图像高频子带系数,抑制图像噪声的同时细匹配图像模糊区域的边缘细节,细匹配图像低频子带.描述图像的模糊熵以及视觉因子,获得待优化的图像子像素细匹配优化目标函数,再依据混合进化算法选取匹配参数,实现低信噪比图像子像素的高精度细匹配.实验结果表明,所提方法能提高图像的视觉效果,具有鲁棒性.