摘要
在深入探讨目前图像分割算法的基础上,针对实际红外图像的分割提出了一种新的方法--动态阈值分割算法。它利用局部空间阈值适度调整全局统计最优阈值,因而其分割阈值是动态的。实验表明该算法能够突出图像中感兴趣的细节,单动态阈值的分割效果相当于常规固定双域值的分割。
On the basis of deeply investigation of image segmentation algorithm, a new methoddynamic threshold segmentation algorithm is proposed against the segmentation of a practical IR image. It moderately adjusts the optimized threshold in global statistics by means of partial spatial threshold, so its segmentation threshold is dynamic. The experiments show that the algorithm can be able to stress the interested details in image. The segmentation effect for the single dynamic threshold is equivalent to the segmentation of the common fixed double-threshold.
出处
《光电工程》
CAS
CSCD
北大核心
2002年第6期57-60,共4页
Opto-Electronic Engineering
关键词
红外图像
动态阈值
图像分割
算法
Image segmentation
Infrared image
Dynamic threshold
