CAJ | 학술논문

针对异源传感器在图像配准时如何精确地找到相应的匹配特征问题,结合红外图像与可见光图像的特征,利用2种传感器对同一场景成像时具有一定的边缘轮廓相似特性,提出了提取图像的Canny边缘,在边缘轮廓上根据边缘曲线上像素点的位置夹角的相似性寻找对应匹配点,通过仿真和实验数据与真实值的比较验证算法的有效性和配准的精度.
침대이원전감기재도상배준시여하정학지조도상응적필배특정문제,결합홍외도상여가견광도상적특정,이용2충전감기대동일장경성상시구유일정적변연륜곽상사특성,제출료제취도상적Canny변연,재변연륜곽상근거변연곡선상상소점적위치협각적상사성심조대응필배점,통과방진화실험수거여진실치적비교험증산법적유효성화배준적정도.
Since the problem of how to accurately find the corresponding matching characteristics by the heterologenetic sensors when image registration, in combination with the characteristics of the red and visible images,the extraction of the Canny edge of images is proposed by utilizing the similar characteristics of the edge outline when the different two sensors image a same scene. That is, find the corresponding matching point at the edge outline according to the similarity of the location included angle between the pixels on the edge curves. The validity and registration accuracy of the algorithm were verified through the simulation and comparison of the experimental data with the true value.