Abstract
Abstract. Digital Surface Models (DSM) derived from stereo-pair satellite images are the main sources for many Geo-Informatics applications like 3D change detection, object classification and recognition. However since occlusion especially in urban scenes result in some deficiencies in the stereo matching phase, these DSMs contain some voids. In order to fill the voids a range of algorithms have been proposed, mainly including interpolation alone or along with auxiliary DSM. In this paper an algorithm for void filling in DSM from stereo satellite images has been developed. Unlike common previous approaches we didn’t use any external DSM to fill the voids. Our proposed algorithm uses only the original images and the unfilled DSM itself. First a neighborhood around every void in the unfilled DSM and its corresponding area in multispectral image is defined. Then it is analysed to extract both spectral and geometric texture and accordingly to assign labels to each cell in the voids. This step contains three phases comprising shadow detection, height thresholding and image segmentation. Thus every cell in void has a label and is filled by the median value of its co-labelled neighbors. The results for datasets from WorldView-2 and IKONOS are shown and discussed.
Highlights
1.1 Problem StatementDigital Surface Models (DSM) which are automatically generated from stereo satellite image matching procedures tend to have many null locations socalled voids or holes
Since the digital elevation models are the basis for all 3D analysis in Photogrammetry and Remote Sensing, i.e. change detection, object recognition and reconstruction, the reliability of the estimated values for these voids is of high importance
In this paper an algorithm for filling voids in DSM derived from stereo satellite images has been developed
Summary
DSMs which are automatically generated from stereo satellite image matching procedures tend to have many null locations socalled voids or holes. These voids are mainly due to mismatching in occluded, shadow and low signal areas. A majority of early reported approaches as well as latter ones include interpolation of holes, since the nature of void filling operation is estimation of unknown values. Simple interpolation sounds efficient to fill the small gaps in DSM, still there are some cases in which it fails In such cases, especially occurring in the automatically generated DSMs from stereo satellite images with a large stereo angle, some holes may cover large areas or even mask some small objects. Amongst all reported methods in this concept, one can find three main approaches including fill and feather (Dowding et al 2004), delta surface fill (Grohman et al, 2006), and moving window applying erosion technique (Karkee et al, 2008)
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
