精确校正图像的径向畸变和倾斜失真

对于用图像进行几何检测,由数码相机生成的图像存在一些固有缺点,如存在径向畸变、倾斜几何变形等。为此提出了一种可同时精确校正数码相机图像的径向畸变和倾斜失真的算法。该算法基于标准网格,首先分别建立倾斜变形模型和径向畸变模型,由此建立目标图像的倾斜变形和径向畸变的联合变形模型;然后用最小二乘法和最优化算法求根据联合失真模型求得标准网格的失真参数,利用该参数可校正同等条件下获得的目标图像;最后对实际图像进行了校正实验,获得了满意的校正结果,说明该算法能有效地同时精确校正图像的径向畸变和倾斜失真。

Precisely Correct Radial and Oblique Distortionof Camera Image

There are radial and oblique distortion in the digital images obtained using the digital camera(DC) in geometric measurement.In this paper,an algorithm to correct the oblique and the radial distortion of the digital image obtained by DC is proposed.Firstly,the oblique distortion model and the radial distortion model of the digital image are established respectively.The union distortion model of the objective image is then derived based on these two individual models.The algorithm to correct the union distortions model of the objective image is designed based on the Least Squared Methods and optimal algorithm,and the distortion parameters of the objective image are obtained from a standard gridding.These distortion parameters are then applied into correcting distortion of the objective image taken in the same condition.The experimental result shows the proposed algorithm could successfully correct the union distortion of the digital image.