摘要
将多个传感器获取的具有不同大小、不同分辨力和不同信噪比的实时图基于Kalm an滤波的方法进行融合,以提高实时图的性能;搜索融合后的实时图在基准图中的位置达到目标定位之目的,即进行景象匹配。在景象匹配过程中,选用归一化互相关系数作为相似性度量。多组实验与分析表明:所介绍的基于多传感器图像融合技术的景象匹配算法可以有效地解决实时图存在部分遮挡、灰度与对比度变化以及复杂噪声干扰等影响下的景象匹配问题。
By fusing the sensed images obtained from multiple sensors with different size, different resolutions and different SNR,tbe performance of the sensed image is improved. By searching the position of the fused sensed image in the reference image,the scene matching is fulfilled. During the process,the normalized cross correlation function is used as the similarity measure. Multiple experiments and theoretical analysis demonstrate that the method is effective even when the sensed images are polluted by partial occlusion, intensity or contrast changes and complex noise corruptions.
出处
《传感器与微系统》
CSCD
北大核心
2006年第9期82-85,共4页
Transducer and Microsystem Technologies
基金
国家自然科学基金资助项目(60234010
60574084
60434020)
国家"973"计划资助项目(2002CB312200)
关键词
图像融合
景象匹配
鲁棒
部分遮挡
灰度
对比度
image fusion
scene matching
robust
partial occlusion
intensity
contrast