An Approach to Detect Vehicles in Multiple Climatic Conditions Using the Corner Point Approach

Journal of Intelligent Systems

Anandhalli Mallikarjun¹  Baligar Vishwanath P.² 

[1] Department of Computer Science Engineering, B.V.B. College of Engineering and Technology, Vidyanagar, Hubballi-580031, Karnataka, India;Professor, Department of Computer Science Engineering, B.V.B. College of Engineering and Technology, Vidyanagar, Hubballi-580031, Karnataka, India;

关键词: vehicle detection;    corners;    grouping;    tracking;   

DOI  :  10.1515/jisys-2016-0073

来源: DOAJ

【 摘 要 】

This paper presents a new method of detecting vehicles by using a simple and effective algorithm. The features of a vehicle are the most important aspects in detection of vehicles. The corner points are considered for the proposed algorithm. A large number of points are densely packed within the area of a vehicle, and the points are calculated by using the Harris corner detector. Making use of the fact that they are densely packed, grouping of these points is carried out. This grouping indicates that the group of corners belongs to each vehicle, and such groupings play a vital role in the algorithm. Once grouping is done, the next step is to eliminate the background noise. The Lucas-Kande algorithm is used to track the extracted corner points. Each corner point of the vehicle is tracked to make the output stable and reliable. The proposed algorithm is new, detect vehicles in multiple conditions, and also works for complex environments.

【 授权许可】

Unknown