General Object Detection Method by On-Board Computer Vision with Artificial Neural Networks

Abstract

The objective of this paper is to find object based solutions for a collision avoidance system. In this paper, the authors present an algorithm for obstacle detection, from the actual video images taken by an on-board camera. The proposed technique is based on Histograms of Oriented Gradient (HOG) to extract features of the objects and classify the obstacles by the Time Delay Neural Network (TDNN). The experimental results showed that it can detect general objects, and is not restricted to vehicles, objects or pedestrians. It has provided good results along with high accuracy and reliability.