Risk detection solution on road based on image processing and deep learning

Abstract

<div lang="en" class="abs"><p class="abs_header">Abstract</p><p>Safety for pupils on the way to school is always the care of their parents. Because of on the way to school pupils can be meet many risks and dangerous issues as accident, violence, kidnapper, and stranger. Thus, their parents always desire to know the status of pupils on the way to school, and they also desire a solution to detect risks on the road and generate warning to pupils. In this paper, we would like to propose a risk detection solution on the road for pupils based on object detection, face detection and distance estimation. The proposed solution uses the techniques of image processing and deep learning to detect dangerous objects, human face and estimate the distance from the detected objects to pupil to give necessary warnings. Experimental results on the road verified that the proposed solution works well, and it have been responded to the purpose of risk detection on the road for pupils in the real.</p></div>