热红外图像与可见光图像的配准与融合  

The Registration and Fusion of Infrared Thermal Image and Visible Light Image

作　　者：郝松傲[1] 秦军[1] 

机构地区：[1]西南交通大学土木学院,四川成都610031

出　　处：《四川测绘》2008年第3期131-133,共3页Surveying and Mapping of Sichuan

摘　　要：红外热成像是一种有效的无损检测技术。红外与可见光传感器是目标识别系统中常用的两种传感器,对这两种传感器图像进行有效地融合能大大提高探测识别的准确性和可靠性。图像配准是图像融合的前提,在仿射变换的模型基础上,用最小二乘法达到用多个配准点拟合出最优参数解的目的。在图像融合中引入了小波变换,实现了图像的彩色融合。Infrared thermal imaging is an advanced NDT means. IR sensor and visual sensor are widely used in target detection and recognition system. The correctness and robustness of target detection and recognition can be enhanced with effective fusion of images from the two kinds of sensors. Image registration is the key premise of image fusion. On the basis of affine transformation model, we use least square method to fit the optimization parameters with multi registration points. Image fusion based on wavelet transformation is presented which is beneficial for accurately carrying out colorful image fusion.

关 键 词：热红外图像 图像数据融合 图像配准 小波变换 

分 类 号：P237[天文地球—摄影测量与遥感]