A Novel Technique for Dead End and Bump Detection in Indian Road Images Using Color Segmentation and Knowledge Base Object Detection

Sudden variation in the surface of the road, sharp turn and irregular road makes it very difficult for safe driving and causes lot of accidents. The situation is more critical especially for vehicles moving with high speed. Sudden bumps or visual dead ends pose a very difficult situation and applying sudden break at those juncture leads to many accidents. So, any technique which can avoid such accidents need to be invented on urgent basis. With this background in view we have proposed to undertake a study by which it will be possible to suggest a novel, efficient technique for detecting the road dead ends, sharp turns and bumps. The technique is such that it segments the road part from the rest of the image, detects the end point of the road by using edge and texture properties and analyzes the end points which are at the far end of the image or much closer. The closer ending of the roads are further thresholded for detecting the dead ends. Bumps are detected by extracting the horizontal profile of the roads and by estimating the volume of horizontal profile. System is tested against the images of the roads taken across different parts of India using digital camera from a moving SUV running at a speed of 30km/h. Experiments show promising results in the detection of both bumps and dead ends.