Automatic Class Mean Calculation of Road Surface from Ikonos Images Using Fuzzy Logic and Particle Swarm Optimization

Automatic road detection from high resolution satellite images has been an active research topic in the past decades. Different solutions are proposed to detect road object such as: fusion-based, fuzzy-based, mathematical morphology, model-based approach, dynamic programming and multi-scale grouping. In this paper, a new fuzzy segmentation method is proposed which is optimized by particle swarm optimization (PSO). The proposed method detects the road network using few samples from its surface. In the IKONOS images, the standard deviation of 10 grey level has been measured for the road classes. In the proposed fuzzy logic system, just one arbitrary pixel up to maximum of three from the road surface is an adequate initial value. The road is identified requiring neither the numbers of the classes nor the corresponding mean values. Particle swarm optimization is used to optimize the proposed fuzzy cost function. The proposed algorithm is applied on real IKONOS satellite image. The results indicate acceptable accuracy for the extracted road surface.