CAJ | 학술논문

图像拼接技术已经广泛应用于各个领域，本文实现了对灰度图像拼接技术，利用角点检测算子来得到图像特征点，采用归一化相关法（NCC）找出匹配角点，通过改进的RANSAC算法消除错误匹配，得到精确度更高的匹配点对，运用加权平均算法进行图像拼接，对拼接后的图像与原图像进行对比，效果很好，具有较高的精度和可靠性，且运行速度良好，可推广应用。
도상병접기술이경엄범응용우각개영역，본문실현료대회도도상병접기술，이용각점검측산자래득도도상특정점，채용귀일화상관법（NCC）조출필배각점，통과개진적RANSAC산법소제착오필배，득도정학도경고적필배점대，운용가권평균산법진행도상병접，대병접후적도상여원도상진행대비，효과흔호，구유교고적정도화가고성，차운행속도량호，가추엄응용。
Image splicing technology has been widely applied in various areas. It use corner detection algorithm based on gray image to obtain the feature point in this paper. The matching corner is found through the normalized correlation method (NCC). To eliminate the error matching as much as possible, RANSAC algorithm has been used. Finally it has gotten the more precise matching point pair using the weighted average algorithm of image fusion. It shows that the method has higher precision and reliability by comparing with the original image to stitching images. And the running speed is fast.