Target Recognition and Intelligent Algorithm Positioning of Mobile Robot with Monocular Vision
Abstract
With the development of science and technology, the autonomous intelligent technology of
mobile robots has developed rapidly. Vision is the main way for mobile robots to obtain
environmental information. Target recognition, as the basis of robot localization, navigation,
and path planning, is a hot topic in mobile robot research. In order to realize the autonomy of
mobile robots, the construction of environmental maps and autonomous positioning are two
basic issues. Vision sensors can provide a wealth of information and have become a common
configuration for mobile robots. It is of great significance to study vision-based mobile robot
target recognition and localization. Based on the above background, the research content of
this article is mobile robot target recognition and intelligent algorithm positioning based on
monocular vision. Aiming at the research background of mobile robot target recognition and
localization based on monocular vision, this paper proposes a mobile robot target recognition
and intelligent algorithm positioning framework based on monocular vision, and designs a
mobile robot using TMS320DM642 and S3C2410 Embedded vision system for target
recognition. Finally, through the experimental simulation and algorithm comparison analysis
of the system, it is concluded that the feature extraction algorithm proposed by PCA-SURF in
this paper has the advantage of small calculation amount. Compared with the monocular
vision SLAM algorithm of fast Harris corner, the PCA-SURF-based single Eye vision SLAM
algorithm has stronger robustness, good real-time processing effect, can accurately locate the
camera, and build a monocular vision SLAM system in real time.