研究目的
Investigating the applicability of the improved algorithm for generating LoS-DSM elevation data through an elevation-based building detection in off-nadir VHR satellite imagery acquired over a dense urban area.
研究成果
The research demonstrates the high success of the disparity-based image-data co-registration and the applicability of the developed LoS-DSM elevations to detecting building objects in off-nadir VHR satellite images acquired over dense urban areas, with an overall detection quality of more than 90%.
研究不足
The generated vegetation index may produce inaccurate results in the cases of roof gardens and buildings with high NDVI values.
1：Experimental Design and Method Selection:
The method involves image pan-sharpening, disparity-based LoS-DSM generation, image segmentation, and building detection and enhancement.
2：Sample Selection and Data Sources:
A subset of VHR multi-view stereo images acquired by the WorldView-2 satellite over a dense urban area in Rio de Janeiro, Brazil.
3：List of Experimental Equipment and Materials:
WorldView-2 satellite images, UNB pan-sharpening technique, multi-resolution image segmentation technique, NDVI for vegetation detection, left-Right visibility check for occlusion detection.
4：Experimental Procedures and Operational Workflow:
The process includes pan-sharpening the reference image, executing the disparity-based LoS-DSM generation algorithm, segmenting the pan-sharpened scene, thresholding off-terrain segments, and enhancing the building map with vegetation and occlusion masks.
5：Data Analysis Methods:
The accuracy of the building detection result was evaluated using typical performance evaluation measures for building detection methods.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
