Camera pose solving method based on color artificial identification

Abstract

The invention provides a camera pose solving method based on color artificial identification, which comprises the following steps of firstly, creating a predefined dictionary automatic generation method to generate a dictionary, and automatically generating a dictionary meeting the requirements according to the number m of the required identifications and the order size n of the identifications; then, realizing the detection of the manual identifier through an image processing technology, and matching the detected identifier with the identifier in the dictionary to complete the identificationof the identifier; and finally, according to the four groups of corresponding points provided by the artificial identifier, solving the planar homography matrix by adopting a levenberg-marquardt algorithm to obtain the pose of the camera. According to the invention, a predefined dictionary automatic generation algorithm is provided, so that the capacity of the dictionary is improved, the coding ofmanual identification is more flexible, and the use scene is expanded, at the same time, the error correction capability of manual identification is improved.

Description

Technical field [0001] The invention belongs to the technical field of artificial recognition, and in particular relates to a method for solving camera pose based on color artificial identification. Background technique [0002] Pose estimation is a research hotspot in the field of computer vision, and it plays an important role in the fields of augmented reality, robot visual navigation, and camera calibration. According to the different ways to solve the pose, it can be divided into hardware-based pose estimation and vision-based pose estimation. The hardware-based pose estimation method can accurately measure the pose of an object through various sensors such as laser, infrared, ultrasound, etc., but the hardware cost is high. Vision-based pose estimation technology can be divided into pose estimation based on natural features and pose estimation based on artificial identification according to different detection targets. The pose estimation method based on natural features ...

AI Technical Summary

Problems solved by technology

The pose estimation method based on natural features solves the pose by detecting key points or textures, but the detection of natural features is difficult, the algorithm has a large amount of calculation, and the efficiency is low, so it is not suitable for applications that require high real-time performance.

Images