Design and recognition of monocular visual artificial landmark based on arc angle information coding

Abstract

According to the requirements of monocular vision positioning of mobile robots in structured environment, a method of artificial landmark design, detection and recognition is proposed to adapt to the large field angle and invariance of rotation. An extensible artificial landmark based on arc-angle information coding is designed, the landmark detection and extraction are realized by structure contour detection, the extracted landmark image is corrected by Hough transform and inverse perspective transformation, After the image filtering and enhancement, the improved circle fitting algorithm is proposed to complete the ring detection and segmentation, and finally the arc angle is calculated by the vector method to identify the landmark and realize the decoding. The experimental results show that compared with the classical QR code method, when using the same imaging system, the landmark design and detection and recognition method proposed in this paper can realize the detection and recognition at longer distances and realize the accurate identification within the range of 0–60 degrees, and has strong anti-noise ability.