Comparative analysis of image processing techniques for obstacle avoidance and path deduction

Abstract

The growing need for automation has a significant impact on our daily lives. Automating the essentials of our society like transportation system has plenty of applications like unmanned ground vehicles in military, wheel chair for disabled, domestic robots, etc., There are driving, braking, obstacle tackling etc., to a transportation system that can be automated. This paper particularly focuses on automating the obstacle avoidance which provides intelligence to the vehicle and ensures a high degree of safety and is performed using image processing algorithms. Edge based detection, image segmentation, and Machine Learning based method are the three image processing techniques used to detect and avoid obstacles. Haar cascade classifier is the machine learning method where Haar cascade analysis is performed for better accurate results with justifying graphs and parametric values obtained. A comparison of the three image processing algorithms is also tabulated considering obstacle size, colour, familiarities and environmental lightings and the best image processing algorithm is inferred.