Abstract: To facilitate computer vision and the application of intelligent robotics in traditional high difference measurements, a high differential transmission measurement method for non-contact application AprilTAG image recognition technology is proposed. As the Robomaster EP robot, the AprilTag label is placed in a known point and the standby point, and the robot camera identifies the AprilTag label on both sides and performs relative position. The three-dimensional sitting is followed, and then the high difference of the rotation center of the robot camera is calculated to be calculated to obtain a high difference from the known point relative to the known point. Through multi-group experiments, the effect of the pitch angle α of the robot camera is studied, and the change of the change of the torch β on the measurement accuracy of the high difference transmission. The experimental results show that when the tension of the robot camera is pitch angle α∈(-2.9°,3.9°], or when the head is angle β∈[-5.5°,1°], AprilTag image recognition is E<0.4, the recognition effect is better, and the two-point high difference error p_x is less than 5 mm, and the accuracy of the high difference transmission measurement can reach millimeters. The above experimental results show that the use of computer vision and intelligent robots have certain feasibility, providing a certain theoretical basis and technical exploration for the R & D industry to develop new intelligent measurement robots in the future surveying industry.

Key words: computer vision, AprilTag image recognition, relative position decisions, intelligent robots, high difference transmission measurement