Non-initial value solving method for three-dimensional spatial similarity transformation model parameters based on linear feature constraints

Abstract

The invention discloses a non-initial value solving method for three-dimensional spatial similarity transformation model parameters based on linear feature constraints, and is applicable to the fieldof three-dimensional spatial imaging. When three or more sets of linear features of the same name exist between two coordinate systems in a known three-dimensional space, one of the coordinate systemsis selected as a reference; with a goal of achieving seamless fusion and unified expression of all spatial three-dimensional data belonging to the two coordinate systems, through model derivation, the specific method and steps of parameter solving are obtained, and further correct and efficient solving of the optimal transformation parameters between another coordinate system and the selected reference system is realized. The steps are simple, the transformation resolution efficiency is high, the conversion parameters of three-dimensional similarity transformation can be calculated without the need to provide the approximate value of the conversion parameters of three-dimensional similarity transformation, the dependence of a classical iterative method on the approximate value of the conversion parameters of three-dimensional similarity transformation is eliminated, and the reliability is high.

Description

technical field [0001] The invention relates to a non-initial-value solving method for three-dimensional space similar transformation model parameters based on linear feature constraints, which is especially suitable for use in coordinate system transformation in three-dimensional space imaging, and belongs to the field of surveying and mapping. Background technique [0002] At present, the three-dimensional spatial similarity transformation plays an important role in the fields of photogrammetric absolute orientation, geodetic coordinate system transformation, computer vision camera pose estimation, and LiDAR point cloud registration. The essence of the three-dimensional spatial similarity transformation is to seek and establish the feature correspondence between the two coordinate systems. Based on the spatial similarity transformation model, such as the Bursa model, the relative position relationship between the two coordinate datums can be described, and the solution is u...

AI Technical Summary

Problems solved by technology

However, the iterative method itself has the following disadvantages: 1) It is necessary to determine the initial value of the transformation parameter in advance, and realize the linearization of the nonlinear transformation model accordingly. Improper selection of the initial value may cause non-convergence of the result; 2) Due to the constraints of the linearization process, this type of method is only applicable to the case of coordinate transformation with small rotation angles. In aspects such as the absolute orientation of photogrammetry, the registration of LiDAR point clouds, etc., if the approximate transformation parameters between the two coordinate systems cannot be provided As an initial value, or the initial value provided is not ideal, the reliability of the calculation result will be seriously affected

Images