Neural Networks for Autonomous Path-Following with an Omnidirectional Image Sensor

Abstract

This paper presents a path-following system implemented with two different types of neural networks, that enables an autonomous mobile robot to return along a previously learned path in a dynamic environment. The path-following is based on data provided by an omnidirectional conical visual system, derived from the COPIS sensor, but with different optical reflective properties. The system uses optical and software processing and a neural network to learn the path, described as a sequence of selected points. In the navigation phase it drives the robot along this learned path. Interesting results have been achieved using low cost equipment. Test and results are presented.

Vision-only Motion Controller for Omni-directional Mobile Robot Navigation

“…An earlier contribution in the area of mobile robot navigation using image features and NN is the work of Pomerleau (Pomerleau, 1989)}. It was then followed by many other research works where they have successively let the robot navigate in the human working environments (Burrascano et al, 2002;Meng & Kak, 1993;Na & Oh, 2004;Rizzi et al, 2002). One very similar research work with the proposed method is the one introduced by Rizzi et al }.…”

An environmental visual features based navigation for mobile robot in a corridor environment

“…It was then followed by many other research works where they have successively let the robot navigate in the human working environments [10][11][12][13].…”

“…One very similar research work with our proposed method is the one introduced by Rizzi et al [10]. The robot in this approach uses an omnidirectional image sensor to grab visual information from the environment and applies an artificial NN to learn the information along the path.…”

Path-following control method for articulated vehicles based on dual heuristic programming

“…In recent years, improving the environmental adaptability and autonomous behaviour of intelligent vehicles has become essential in academia. Intelligent control-based methods, 12,13 such as fuzzy-based control 14,15 and neural network methods 16,17 have emerged. Taghavifar et al 18 used an extended Kalman filtered fuzzy neural network observer and a type-2 fuzzy PID neural network to develop an improved fuzzy type reduction strategy and an adaptive compensator to eliminate lateral deviations.…”