摘要
针对同一场景的红外和可见光图像,提出一种基于OTSU递归分割算法的区域分割和非采样Contourlet变换的红外和可见光图像融合算法。首先,对红外和可见光图像进行区域分割和区域关联,关联映射图分为目标区域、背景区域和灰度区域。随后,应用NSCT变换对图像进行多尺度、多方向分解,按照关联映射图中三个区域对NSCT分解后的高低频子带系数进行区域划分,根据不同区域的特性在NSCT域设计不同的融合规则。最后,进行重构得到融合图像。对三组不同场景图像的实验结果进行主观目视判别和客观性能评价,对比基于像素和邻域能量的融合算法,本文算法不仅能较全面的保持可见光图像中的光谱信息,而且能够有效、准确的提取红外图像的热目标信息,优于传统的基于像素和邻域能量的融合算法,可获得较理想的融合图像。
Aimed at the infrared and visible light images in the same scene,a fusion algorithm based on region segmentation and nonsubsampled contourlet transform(NSCT)is proposed in this paper.Firstly,regional segmentation and regional association are used in the infrared and visible light images,the joint region map is divided into the target area,the background area and gray area.Then,the source images are performed to multi-scale and multi-direction decomposition in NSCT domain,after which both low-pass sub band coefficients and band-pass directional sub band coefficients are divided into three areas in accordance with the joint region map.According to the characters of the different areas,different fusion rules are designed in NSCT domain.Finally,the fused result is obtained through inverse NSCT.The propose algorithm has been experimented on three different scene images,experimental results are compared both in subjective and objective standards.It is showed that the algorithm can not only keeps the spectrum information of visible light image completely and richly,but also extracts the target characters of infrared ray image accurately and effectively.The proposed algorithm is superior to those conventional fusion methods based on nonsubsampled contourlet transform,using pixel or neighborhood energy,and is feasible and effective,also can manifest better fusion effect.
出处
《激光与红外》
CAS
CSCD
北大核心
2010年第11期1250-1257,共8页
Laser & Infrared
基金
国家高技术研究发展计划("863"计划)(No.2005AA778032)资助