Rule-Based Multiple Target Monitoring in Robot Vision Using Partitioning

Abstract

© Published under licence by IOP Publishing Ltd. In this study, the researchers propose an algorithm and implement it as a robotcamera which not only detects many moving objects and captures them, but also follows the moving objects to continue capturing and monitoring. The robot uses Gaussian filtering to detect all motions and then applies partitioning on the scene to obtain location of all moving objects in three major areas and different sub-areas. Based on the partitioning and position of the moving objects, the robot might be in any of unsafe state, safe state, and over-safe state. Regarding each state, the robot performs proper movement following two cardinal rules to pursue monitoring the moving objects.

Source or Periodical Title

Journal of Physics: Conference Series

ISSN

17426588

Document Type

Article

This document is currently not available here.

Share

COinS