A QR Code Image Processing Mechanism for Book Access Robot Positioning and Navigation

摘要

In The mapping system of current libraries is based mainly on the inertial navigation and binocular visual navigation technology. Due to the limited mapping accuracy, however, such a system cannot meet the high precision navigation, positioning and book information recognition requirements of the library automatic access robots. To improve the robot system locating accuracy, we propose a QR code image processing mechanism for library robot positioning and navigation system based on the visual technology. Specifically, with the binocular cameras that are fixed on robots to quickly establish the three-dimensional coordinates of the QR code, such a mechanism tries to extract the Haar features of the QR code images and identify the target area quickly with the Adaboost algorithm, and then it controls the robot arms with code reader to access the accurate information of the QR codes, and finally the realize accurate book information reading and navigation accuracy improvement functions with an information correction function. Experiments are also conducted to verify the effectiveness of the proposed mechanism.