A method of visual measurement based on QR code in navigation and positioning of closed-type orchard

Abstract

To realize the navigation and positioning of agricultural robot in closed-type orchard, this paper presented a method of visual measurement based on QR code. Firstly, we constructed a monocular vision measurement model based on the QR code target, then extracted the structural feature points of the QR code target in the orchard by image processing, and combined camera calibration and vision measurement gaining the relative pose between the camera and the QR code target. According to the results of experiments, this method boasts good real-time performance as the longest consuming time of applying this method to process an image for visual measurement is 0.23ms, in addition, the repeatability accuracy tests show that the maximum translation deviation is less than 3.2cm and the maximum rotation deviation is less than 2.9 ° in the range of measurement.