一种基于抛物面模型的参数修正的鱼眼图像校正算法OA

Fisheye lenses,renowned for their ultra-wide field of view,find extensive applications in security surveillance and exploration fields.However,radial distortion caused by the short focal length design impairs the practical utility of fisheye lenses seriously.In view of the above,this paper introduces a parameter-adjustable fisheye image correction algorithm based on a paraboloid model,which enhances the correction quality of fisheye image effectively.Initially,an improved row-by-row and column-by-column scanning method is employed for effective region extraction,which reduces scanning time substantially.Subsequently,a preliminary correction is performed on the fisheye images by a parameter-adjustable paraboloid model,and then the model parameters are adjusted according to the correction effects for a subsequent refinement.Finally,a bilinear interpolation algorithm is applied to the corrected images for interpolation.Simulation results demonstrate that,when compared with perspective projection algorithm featuring optical center shift and longitude correction algorithm involving repositioning,the proposed algorithm yields superior correction effects in outdoor architecture,indoor structures,and landscapes,preserving more valid information in the corrected images.This makes it particularly valuable for practical applications in the fields of security surveillance and exploration.