Road Extraction and Car Detection from Aerial Image Using Intensity and Color

Abstract

In this paper a new automatic approach to road extraction from aerial images is proposed. The initialization strategies are based on the intensity, color, and Hough transform. After road elements extraction, chain codes are calculated. In the last step, using shadow, cars on the roads are detected. We implemented our method on the 25 images from "Google Earth" database. The experiments show an increase in both the completeness and the quality indexes for the extracted road.

Keywords

• Road extraction
• car detection
• machine vision
• hough transform
• chain code