摘要
针对镜头径向失真的问题,提出了一种基于除法模型(DM,division model)和核密度估计算法的校正方法。该方法利用同一场景两个不同视点的失真图像对建立DM模型。首先对失真图像对进行特征点的提取和匹配。然后根据基本方程式联立方程组,采用核密度估计算法求出最优解即为校正参数。最后根据求得的校正参数对两幅失真图像进行校正。仿真表明,核密度估计算法提高了失真校正的鲁棒性和图像校正的精度。
We present a radial distortion correction method based on division model and kernel density estimator algorithms for radial lens distortion problems. This method uses two distorted images from two dif- ferent viewpoints of the same scene to create the DM model. First, we use the DM radial distortion model to select and match feature points of the two distorted images. Second, we select feature points to couple the simultaneous equations group. The root of equations is obtained by use of kernel density estimation al- gorithm and the best root is the calibration parameter. Finally we correct the two images based on the solved calibration parameter. The simulation results show that the kernel density estimation algorithm improves the accuracy of images and robustness of the lens correction.
出处
《南京邮电大学学报（自然科学版）》
2011年第6期32-36,43,共6页
Journal of Nanjing University of Posts and Telecommunications：Natural Science Edition
基金
国家自然科学基金(61071091)
国家自然科学基金青年科学基金(61001152)
南京邮电大学校科研基金(NY210053)资助项目
关键词
径向失真
除法模型
核密度估计算法
基本方程式
失真校正
radial distortion
division model
kernel density estimation
basic equations
correct distortion
