Target multi-station attitude processing method based on multi-linear vector mapping matching

Abstract

The invention provides a target multi-station attitude processing method based on multi-linear vector mapping matching, and belongs to the field of conventional target range optical attitude measurement. For axial symmetry revolving body target attitude processing, the method comprises the following steps: firstly, determining object characteristics of a middle axis or parallel vector, and then extracting image characteristics of the middle axis or parallel vector; and finally solving the yaw angle and the pitch angle by establishing a linear feature object image mapping relation. For non-axisymmetric revolving body target attitude processing, firstly, multiple uncorrelated straight line object features are determined, then corresponding straight line image features are extracted, and yaw, pitch and roll angles are finally solved by establishing a straight line feature object image mapping relation. According to the method, on the basis of object image space corresponding mapping of necessary linear vectors, the three-attitude-angle measurement precision and stability are guaranteed, meanwhile, the effective shielding resistance is high, and the method can be widely applied to the field of rigid body type target attitude processing with key linear features.

Description

technical field [0001] The invention relates to the field of conventional weapon testing and testing, in particular to the field of shooting range optical attitude measurement. Background technique [0002] There are currently three processing methods for the medium and long-term attitude processing of typical targets in the shooting range, namely the absolute orientation method based on multiple (more than 3) corner points with the same name, the processing method based on target model projection matching, and the surface-to-plane intersection method based on the central axis . Among the above three methods, the absolute orientation method based on the corner points with the same name is only applicable to the case where (more than 3) corner points with the same name can be extracted synchronously, and the processing principle must be based on the reliable spatial intersection of feature points, and further rotation to obtain Attitude parameters have higher requirements fo...

AI Technical Summary

Problems solved by technology

The method based on the projection matching of the target model needs to establish a relatively accurate model in advance. In addition, the model matching method involves projection matching of a large number of points, and the calculation amount is relatively large, time-consuming and inefficient.

The face-to-face intersection method based on the central axis is more advantageous in the feature extraction link, and it does not need to match the same name, and it is more flexible in feature extraction. It only needs parallel sub-station vectors, but its disadvantages are that there is no redundancy in information and no feedback optimization. link, the accuracy depends entirely on the non-redundant interpretation accuracy and the reliability of the intersection interval; in addition, only the yaw angle and pitch angle information can be processed

Images