Adaptive non-parametric identification of dense areas using cell phone records for urban analysis
Engineering Applications of Artificial Intelligence
| Hi-index | 0.00 |
Automatic building extraction is a bottleneck of automatic surveying and mapping of big scale urban image. For the single aerial gray-scale image, the automatic buildings extraction, an algorithm based on the principle of confrontation and using the priori knowledge of building is presented here. A two-dimension OTSU segmentation will be used to do the rough segmentation because of the characteristics of high-brightness which the roof of the building has; then we use mathematical morphology, image segmentation, edge detection and other measures to detect the non-building, so using the principle of confrontation and removing the non-buildings, the buildings can be extracted. At last the extraction result will be verified using the information of the building shadow, the area of building roof, altitude information and other prior knowledge, in order to improve the accuracy of building extraction in the aerial image. Moreover, the method proposed is test in our experiment using the true-color aerial image. Theoretical analysis and experimental results show building can be extracted effectively by means of the algorithm proposed and it is necessary to research deeply.