Using LIDAR data and airborne spectral images for urban land cover classification based on fuzzy set method

In this paper, we propose an analysis on the combinative effect of high-resolution airborne image and light detection and ranging (LIDAR) data for the classification of complex urban areas. In greater detail, the proposed system is composed of three models briefly. Model one includes an advanced kernelized fuzzy c-means classification method for high-resolution airborne image. The characteristics of LIDAR point cloud are introduced in model two, membership degree function of buildings, vegetations and naked land have been built. In model three, high-resolution image and elevation data form LIDAR point cloud are jointed. Experiment carried out on a complex urban area provide interesting conclusions on the effectiveness and protentialities of the joint use of high-resolution image and LIDAR data. In particular, the elevation data was very effective for the separation of species with similar spectral signatures but different elevation information. Experimental results approve that elevation data can improve classification accuracy in building occupied area obviously.