期刊名称:ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN:2194-9042
电子版ISSN:2194-9050
出版年度:2006
卷号:XXXVI Part 3
出版社:Copernicus Publications
摘要:Automated image orientation is still a key problem in close-range photogrammetry, in particular if wide baseline images are employed. Nowadays, within the image-based modeling pipeline, the orientation step is the one which could be fully and reliably automated, exploiting the potentiality of computer and image processing algorithms. In this paper, we summarize recent developments in this field and apply them in three different workflows to automatically extract markerless tie points from close- range images of different types (video sequence, large and wide baseline images). Furthermore we compare the results obtained from bundle block adjustment using the automatic tie points with the results obtained by manual measurements and show how the accuracies of the automatic tie point extraction can further be improved by including least squares matching techniques