结合Top-hat变换改善同态滤波的方法

Combining with Top-hat transformation to improve homomorphic filtering

同态滤波是处理照度不均匀引起图像质量下降的有效方法,可以平滑背景,提升对比度,增强图像细节.在实际使用中,每个图像成像的物理条件都不尽相同,同态滤波最佳参数的选择,需要经过大量实验确定,否则效果不佳.Top-hat变换有助于平滑背景,增强图像阴影部分的细节.提出了结合Top-hat变换改善同态滤波的方法,以减少因滤波参数选取所带来的开销.使用Matlab软件进行仿真,通过图像三维可视化效果、均值、标准差、信息熵、图像对比度客观评价处理后的图像.实验表明,此方法不仅降低了同态滤波对参数的依赖同时也增强了滤波效果.

Homomorphic filtering method is usually used for uneven illumination image enhancement,since it can smooth background,enhance contrast,and enhance image details.While the diversity of physical conditions of image imaging leads to difficulty of the selection of optimal parameters for homomorphic filtering,which usually needs to be determined through experiments for the better enhancement effect.Top-hat transform helps to smooth background,enhance image shadow part details.An improved method of homomorphic filtering combining with the top-hat transform is presented to reduce the overhead caused by the filter parameters selection.The performances of given method are evaluated on the simulation of Matlab software.The enhancement results are analyzed by image three dimensional visualization effects,mean value,standard deviation,information entropy and image contrast value.Experiments show that our method is feasible and effective for reducing dependence of homomorphic filtering on filtering parameters and enhancing the filter effect as well.