Road Terrain detection and Classification algorithm based on the Color Feature extraction

摘要

Today, In the content of road vehicles, intelligent systems and autonomous vehicles, one of the important problem that should be solved is Road Terrain Classification that improves driving safety and comfort passengers. There are many studies in this area that improved the accuracy of classification. An improved classification method using color feature extraction is proposed in this paper. Color Feature of images is used to classify The Road Terrain Type and then a Neural Network (NN) is used to classify the Color Features extracted from images. Proposed idea is to identify road by processing digital images taken from the roads with a camera installed on a car. Asphalt, Grass, Dirt and Rocky are four types of terrain that identified in this study.