论文部分内容阅读
内定向是摄影测量的一个基本问题,其中框标位置的近似确定是最重要也是最困难的工作。当框标不是对称形式且淹没在背景当中,框标与影像叠置在一起,传统的灰度匹配方法已不再适用,文章主要研究此种情况下的自动内定向,采用影像二值化、区域生长法分割影像、边界跟踪将框标所处的线特征提取出来,进行线特征匹配,最终实现框标的自动识别与定位,其自动内定向成功率达到了98%以上,在实际生产中得到推广与应用。
Internal orientation is a basic problem of photogrammetry, and the approximate determination of the position of the frame mark is the most important and most difficult task. When the frame standard is not symmetrical and submerged in the background, the frame standard overlaps with the image, the traditional gray matching method is no longer applicable. In this paper, we mainly study the automatic internal orientation in this case, using image binarization, Regional growth segmentation method, boundary tracking will be extracted from the line features of the box mark, the line feature matching, and ultimately the automatic identification and positioning of the frame mark, the automatic internal orientation success rate of more than 98% obtained in the actual production Promotion and application.