基于熵空间理论和小波变换的红外图像纹理分割

摘要

The jamming of the background to the target is an important factor affecting the infrared equipment operational range. The research on target recognition method of the infrared image whose background is fog has important implications on the application of infrared equipment. This paper has broken through the traditional recognition method for infrared image target. Through research on the fog texture, one kind of natural texture, we extract texture energy of the infrared image on several frequency bands by continuous wavelet transform(CWT), and we have realized texture segmentation of the infrared image whose background is fog based on decomposition and synthesis theory of different particle size space of entropy space.%　　背景对目标的干扰是影响红外装备作用距离的重要因素，研究以雾为背景的红外图像中目标的识别方法对红外装备的应用有着重要的意义。突破了传统的红外图像的目标识别方法，通过对雾这种自然纹理的研究，采用小波变换的方法对红外图像提取各个频道上的纹理能量，基于熵空间理论的不同粒度空间分解与合成思想有效实现了以雾为背景的红外图像的纹理分割。