Relative orientation optimization method based on bound constraint function

Abstract

The invention discloses a relative orientation optimization method based on a bound constraint function; the method comprises the following steps: creating the bound constraint function; creating a relative orientation optimization equation containing the bound constraint for n pairs of corresponding image points; converting the relative orientation optimization equation containing the bound constraint into an unconstrained minimization equation; using a least square generalized inverse method to obtain the unconstrained minimization equation approximation optimal solution; terminating iteration when terminal criterions are met, and using the finally solved relative orientation element as the relative orientation optimization equation (containing the boundary constraint ) approximation optimal solution. The method creates the bound constraint function, stably limits the relative orientation element optimal solution in a feasible domain, thus preventing optimization solving failure caused by the iteration point crossing obstacle and disengaging from the feasible domain; in addition, the method uses a sharpness factor to control the bound constraint function climbing speed outside the feasible domain, uses indexes to control the penalty factor decreasing speed, thus fast and reliably converging the relative orientation element to the optimal solution.

Description

technical field [0001] The invention belongs to the technical field of machine vision and photogrammetry, and in particular relates to a relative orientation optimization method based on bounded constraint functions. Background technique [0002] With the continuous improvement of the performance of SLR cameras, the use of SLR cameras in free photography mode does not require precision calibration plates and high-precision camera motion platforms with high processing and maintenance costs to achieve self-calibration of cameras, and to acquire images quickly, flexibly and at low cost. Measuring the three-dimensional data of industrial site parts has become one of the development trends of coordinate measurement technology. [0003] By solving the coplanar conditional equation of the same-named image point, the relative orientation element is obtained, which is the basis of the three-dimensional measurement and camera self-calibration based on the SLR camera, and is also the b...

AI Technical Summary

Problems solved by technology

[0015] In order to overcome the above-mentioned shortcomings of the prior art, the present invention proposes a relative orientation optimization method based on bounded constraint functions, which determines the constraint interval according to the initial value accuracy of the relative orientation elements, and obtains the feasible region for solving the coplanar equation; on the other hand, Construct a unique bound constraint function to limit the optimal solution within the feasible region, ensuring that the optimal solution of the relative orientation elements can be reliably and quickly converged even when the initial value accuracy is not high

Images