Moving Car Detection using HOG features


Patel Krupaben Pravinkumar , Sankalchand Patel College of Engineering; Manish I. Patel, Sankalchand Patel College of Engineering


Car Detection, Histogram of Oriented Gradients, Support Vector Machine, Hard Negative Mining


Vision is the most powerful sense of the five human senses. This paper proposes real-time monocular-vision based techniques for the multiple car detection. Car detection system should minimize driving difficulties in various conditions, like in sleeping, busy in communication devices etc., thereby reducing traffic accidents. Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) are used in this work. SVM is the classification method. HOG will extract the features of given frames extracted by video and that important features will be given to the SVM. Before that we have to give training dataset to SVM to train it, because SVM is supervised learning algorithm. According to training it will classify data and accordingly present or absent of car is decided. Simulation results are presented for different images, which show accuracy of suggested approach.

Other Details

Paper ID: IJSRDV4I30001
Published in: Volume : 4, Issue : 3
Publication Date: 01/06/2016
Page(s): 250-254

Article Preview

Download Article