复杂背景的红外图像过渡区提取与分割

Transition Region Extraction and Segmentation of Infrared Image on Complex Background

在基于个数判断噪声的中值滤波器进行平滑处理的基础上,提出了一种改进的过渡区提取方法,并将该方法应用于对比度低、宽谷直方图的复杂背景红外图像的分割。Matlab仿真实验表明,该方法减少了图像分割的运算量,并能将具有复杂背景的红外图像目标分割得较好。

A median filter based on estimation of noise is designed for smoothing processing, and put forward an improved transition region extraction method. Finally, this method is used to deal with segmentation problems for those infrared images whose peak values of histogram have great difference. Matlab simulation experiments showed that the proposed method can decrease the time of calculation and provide satisfactory segmentation for those infrared images on complex background.