Color stereoscopic calibration object-based unmanned aerial vehicle calibration method and system

Abstract

The invention discloses a color stereoscopic calibration object-based unmanned aerial vehicle calibration method and system. The method comprises the steps of putting a color checkerboard stereoscopic calibration object into a to-be-shot scene; shooting an image of the color checkerboard stereoscopic calibration object in at least three different directions by adopting an unmanned aerial vehicle; linearly solving intrinsic parameters of an unmanned aerial vehicle camera by adopting a vanishing point theory according to the shot image of the color checkerboard stereoscopic calibration object; and determining a geometric restriction relationship between the spatial position of the unmanned aerial vehicle camera and the image by adopting a coordinate projection transformation method according to the intrinsic parameters of the unmanned aerial vehicle camera. According to the method and the system, the color checkerboard stereoscopic calibration object is adopted for performing camera calibration, so that accurate measurement is facilitated, the detection precision is high, the placement is facilitated, and the universality is high; and the calibration of the intrinsic parameters of the camera can be finished only by shooting the image of the color checkerboard stereoscopic calibration object in the at least three different directions and obtaining the intrinsic parameters of the unmanned aerial vehicle camera in combination with the vanishing point theory, so that the use is more convenient. The method and the system can be widely applied to the field of computer vision.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to an unmanned aerial vehicle calibration method and system based on a color stereo calibration object. Background technique [0002] The ultimate goal in the field of computer vision research is to enable computers to have the same ability to perceive the 3D environment as humans. In this field, the image-based 3D reconstruction process is generally the reverse process from 2D image restoration to 3D scenes, that is, through computer analysis. The two-dimensional images of the target scene or object under different viewing angles restore the three-dimensional spatial geometric information of the scene or object. 3D reconstruction has a wide range of applications and plays a huge role in a variety of scenarios, such as the restoration of precious historical relics and ancient buildings, 3D reconstruction of medical human organs, autonomous driving of unmanned vehicles, etc., using ord...

AI Technical Summary

Problems solved by technology

[0006] 3D reconstruction technology generally needs to take images of the target object from multiple angles through the camera. If the distance between the camera and the object is fixed, the spatial range that can be reconstructed is very limited

Since the marker points and reference objects are uncertain, surveying and mapping is required at any time, the measurement accuracy is not easy to guarantee, and the actual operation is cumbersome, and there are many uncertain factors. If high-precision 3D reconstruction is to be achieved, it is necessary to set up calibration objects

If simply adopting the planar black-and-white checkerboard calibration object in the existing ground close-range photographic exploration method, it is difficult to provide three-dimensional information; if a combination of two checkerboard calibration planes is used, it cannot meet the needs of observation and calibration at any position in space; and If the black and white checkerboard is used on multiple surfaces of the cube at the same time, there will be certain visual recognition difficulties and large errors for both photographers and photo users

Images