OBSTACLE DETECTING ROBOT FOR UBIQUITOUS ENVIRONMENT - DEVOID OF SENSOR

Abstract

Autonomous path tracking robot without any knowledge about the environment must muddle through the
obstacle. The onboard sensing technique uses sensor to detect obstacle. The vision based technology progress with
camera to become aware of obstacle. On comparing the above two approach, the latest one provides detailed
information about the environment which is not obtained by using sensor. There are some drawbacks in onboard
sensing technology like selection of sensor based on color and lightning condition which can be overcome by the
proposed method. The proposed method depends on vision based technology which is capable of providing detailed
description about the environment to avoid obstacles. The piecewise linear contrast stretching is used for image
segmentation. The map less navigation where the robot is not trained to any environment behaves like a human
when the robot detects an obstacle from the captured image. Once the obstacle is detected the robot takes another
possible path for the further safe navigation.