摘要
针对红外图像的特点,提出了一种改进的红外图像分割算法。该算法综合应用模糊熵、遗传算法和数学形态学等技术,从算法流程上对传统的分割算法进行了改进。仿真结果表明,文中提出的算法能快速准确地实现红外图像分割,鲁棒性好,自适应性强,能够针对不同的复杂地面背景,给出较为理想的分割效果:克服了传统分割算法的过分割和欠分割问题。
An improved infared image segmentation algorithm is proposed in view of the features of infared image. The algorithm, adopting the technique of fuzzy entropy, genetic algorithm morphology and etc, improves the traditional segmentation algorithm in the algorithm flow. The simulation results shows that the proposed algorithm is fast and reliable for infared image segmetation. And the algorithm, having good robust and adaptive, could give a ideal segmentation result aiming at different and complicated ground background, and conquer the the over-segmentation and less-segmentation problems of the traditional segmentation algorithms.
出处
《激光与红外》
CAS
CSCD
北大核心
2007年第6期579-582,共4页
Laser & Infrared
关键词
红外图像
分割算法
模糊熵
遗传算法
形态学
infared image
segmentation algorithm
fuzzy entropy
genetic algorithm
morphology
