红外图像目标分割方法研究

Research on the Target Segmention Method of Infrared Image

针对红外图像目标准确分割的难题 ,提出了一种有效的目标分割方法。该方法所遵循的基本准则是 ,使区域内部所考虑的特征或属性相同或相似 ,而这些特征或属性在不同区域中则是不同的或存在差异的。依据这个准则运用最大距离法和自动增强图像分割门限的方法实现了红外图像中目标的准确分割 ,并且通过实验验证了本方法的有效性。

With the view of segmenting a target accurately, this paper describes a target segmentation algorithm. It is based on the following criterion, i.e., the features or property are the same or similar within the regions, while different in different regions. According to this criterion,we used the maximum distance and automatic enhanced image segmentation threshold methods to realize an accurate infrared image target segmentation . The experimental result shows that the method is practical.;