Extraction And Application Of Textured Vertical Facades From Aerial Imagery

The simulation of three-dimensional scene, especially infrared simulation iswidely applied in performance evaluation of the imaging system and battlefieldsimulation, etc. Texture is an important resource for the scene to embody the detailand realism of the model, the massive collection of textures is taken more and moreattention. With the rapid development of remote sensing technology, the resolution ofimage get more accurate, therefore, aerial images can express the real target featuredetails clearly.Based on the aerial images, the data of facade texture of buildings is extractedand the texture library automatically is generated. Firstly, adaptive Canny edgeoperator is used for edge detection of images. The direction of vertical edges areobtained based on Vanishing Point, and the direction lines are detect by pixel trackingalgorithm. Secondly, after preprocessing the image, improved straight line extractionalgorithm is used for edge extraction. Based on perceptual organization, the detectionand extraction of facade faces texture are achieved. Finally, the detailed stepinstructions are given for the problems of texture deformation, as well as the texturequality control and coding. The buildings are reconstructed by two applicationalgorithm.The experimental results show that this approach is effective.