An Active Pose Relocalization Method Without Camera Calibration

Abstract

An active pose quick relocation method without camera calibration comprises the steps of judging whether a relocation standard is met or not, if the relocation standard is met, finishing the flow; otherwise, entering the next step: calculating a homograph transformation matrix H, namely extracting matched feature point pairs of a current image and a reference image, and calculating the homograph transformation matrix H according to the extracted feature point pairs; dynamically updating an indicating box according to the solved homograph matrix H; according to the dynamically updated indicating box, obtaining an indicator for adjusting and controlling six degree of freedom motion of a camera, wherein the six degrees of freedom comprise rising / dropping, leftward turning / rightward turning, leftward swing / rightward swing, leftward moving / rightward moving, upward moving / downward moving, forward moving / backward moving; performing camera pose adjustment according to the indicator for adjusting and controlling the six degree of freedom motion of the camera; repeating the steps till the relocation standard is met, and finishing. The method provided by the invention is good in real time performance, high in flexibility, high in location accuracy, not sensitive to light conditions and suitable for various work environments indoors and outdoors.

Description

technical field [0001] The invention relates to a fast repositioning method for poses. In particular, it involves an active pose fast relocalization method without camera calibration. Background technique [0002] (1) SIFT: short for Scale Invariant Feature Transform, an algorithm in computer vision. It is used to detect and describe local features in images. It looks for extreme points in the spatial scale and extracts its position, scale, rotation invariant. The description and detection of local image features can help to identify objects. The SIFT feature is based on some local appearance points of interest on the object and has nothing to do with the size and rotation of the image. The tolerance to light, noise, and some viewing angle changes is also quite high. For these properties, they are highly significant and relatively easy to obtain. Using SIFT features to describe the detection rate of partial object occlusion is also quite high, and even more than 3 SIFT ...

AI Technical Summary

Problems solved by technology

[0008] Although the above two technologies are relatively mature, they are only applied to image processing and cannot solve the problem of active pose relocation

Images