A New Approach to Images’ Pespective Transformation in Robotics Telemetry

Abstract

Remotely controlled robots are the majority of contemporary robots’ population. As a rule such robots are used for inspection, patrolling, mines disposal. This is caused by a fact that today’s level of robots’ autonomy is rather low and autonomous robots can not secure reliable performance. However performance of remotely operated robots depends largely on efficiency of information perception by human-operator. This paper studies images representation to operators on robots’ control pendants. More specifically it studies 3D images representation on flat displays. The goal of this study is to increased reaction rate and decrease errors done by robot’s operator due to image’s uncertainty or poor quality. Our main attention is paid to the development of software that implements the technology of transforming the perspective of images. Matrixes for perspective transformation are studied. Hence application of these matrixes within remotely operated robots is discussed. A novelty of this research is in new knowledge about perspective transformation which is done for a better information perception by human-operator and, as an outcome, to increase efficiency of remotely operated robots. The related technologies in the field of telemetry and technical vision systems have been investigated. Also considered are works in medical fields, in particular, the psychology of perception of images and space. A static software model has been developed. The video camera has been implemented with the introduction of perspective distortions to improve the reliability of the transmission of the necessary areas in the image. Research on the technology is carried out jointly with the Institute of Biomedical Problems of the Russian Academy of Sciences. A special prototype for perspective transformation basing on various scenarios is being developed.