Road Segmentation in Street View Images Using Texture Information

摘要: Road segmentation is a problem encountered fairly frequently, especially in the framework of scene understanding and self-driving cars. On flip side, there are several Street View databases that offer large amounts useful data, which still relatively untapped. In this paper we propose road algorithm specifically aimed at segmenting roads from street view images. We use fisher vectors to encode small windows extracted main image multiple scales, then classify these patches voting scheme get final segmentation. optionally utilize spatial prior superpixels improve our Our performs well outputs good for further evaluation. test method on KITTI dataset compare it state art.