URBAN LAND COVER CLASSIFICATION USING AERIAL LIDAR AND CCD IMAGES

Timely and accurate acquisition of land surface information in urban area is essential in monitoring the process of urbanization, it is critical for urban planning and management. Aims to investigate urban land cover classification using high resolution remote sensing data, this study proposed an object-oriented image analysis based approach for urban land cover mapping using LiDAR and CCD images. The proposed approach based on the analytic hierarchy process, the optimal scale for image segmentation and methods of image classification are investigated. The main land cover types are determined as low built-up, high built-up, tree, grassland and road, with the multi-scale image segmentation, the corresponding feature spaces are adopted for fuzzy classification. The experiment was carried out in Zhangye city, Gansu province. The results show that the overall accuracy is 93.42, and the Kappa is 0.91.