基于角点的红外与可见光图像自动配准方法

针对红外图像与可见光图像的自动配准问题,提出了一种基于图像角点特征以及仿射变换模型的方法.利用Harris因子分别在红外图像和可见光图像上检测角点,并对两幅图像进行边缘检测,得到其边缘图像.通过角点邻域在边缘图像上的相关性,实现角点的粗匹配;通过角点的细匹配,从匹配的角点中选择两对匹配最佳的点作为仿射变换的控制点,得到仿射变换模型,并对待配准图像进行仿射变换,从而实现图像配准.实验结果表明:该方法运算速度快,可以很好地完成红外与可见光图像的自动配准.

Infrared-Visual Images Automatic Registration Based on Corners

A new method based on the corner feature and affine transformation is presented in order to realize automatic registration of infrared image and visual image,including corner extraction,edge extraction,corner match and affine transformation.Harris factor is used to detect the corners,and Canny operate is used to extract the edges in two images.The correlation of corner neighborhood is used in edge images to realize coarse corners match.Corners match is refined,and two pairs of corners are chosen as reference point of affine transformation,which are the best in matched corners.The affine transformation model is obtained,and affine transformation in the image prepared for registration is carried out.The experiment results show that the method can well complete automatic registration of infrared image and visual image at a faster registration speed.