基于最大累积剩余熵的红外图像分割

摘要

经典阈值化分割方法是常用的图像分割方法之一,但对于低对比度,低信噪比的红外图像分割效果不理想,将累积剩余熵运用于红外图像分割,即寻求使得累积剩余熵最大的灰度值作为分割阈值,实验结果表明:基于最大累积剩余熵的图像分割方法是一种有效的红外图像分割方法,比起经典的阈值分割方法,分割红外图像的效果要好.%Classical thresholding method is one of the frequently used image segmentation method,but it can not obtain good segmentation result for infrared image with low contrast and low SNR. The cumulative residual entropy is applied to the infrared image segmentation. The method seeks gray value which make the maximum cumulative residual entropy as the threshold value. Experimental results show that the maximum cumulative residual entropy is an effective infrared image segmentation method,and can get better segmentation result for infrared image than the classical thresholding method.