Acquisition of Object Pose from Barcode for Robot Manipulation

摘要

General, robots obtain poses of target objects by matching the images of the observed objects with data in database. However, the process of matching images costs so long time that robot's action become slow. In order to shorten response time for robots searching target objects, we propose a method for robots to obtain information of poses of observed objects by calculating corner points of barcodes on the objects. Since information in a barcode is less than the one in an image, the method can help robot rapidly obtain the information of its target objects. Furthermore, in order to reuse the method in other robot systems, we create a RT-Component(RTC) to realize the method.