Region adaptive correction method for radial distortion of fish-eye image

A fisheye image barrel distortion correction method of the straight slope constraint

“…In the paper, the mean square deviation of the corresponding homologous image point residual errors of the correction image and the ideal image can be regarded as indicators to evaluate distortion correction precision in [17], (10) where, represents the node on the row and the column , is the corrected imaging position, is the ideal imaging position. The mean square deviation of the residual error is less, the correction effect is better.…”

“…In [7][8][9][10], the constraint that the projection of a straight line is still a straight line is utilized for gaining the distortion coefficients. Although the division model [13][14] established by the method is simple, the correction precision is poor.…”

“…Therefore, we present a fisheye image barrel distortion correction method based on the straight slope constraint in the paper. The algorithm not only can be used for eliminating the residual errors on the edge of the image and asymmetry, but also can improve segmentation effects of the high dynamic fisheye image compared with [6][7][8][9][10][11][12]. Firstly, a six-parameter radial distortion polynomial is established according to the distortion model proposed by Slama [15], Tasi [16], etc.…”

“…Methods based on a single image in [6][7][8][9][10][11][12] refer that distortion parameters are determined by the linear constraint. The methods are flexible, and only a part of camera parameters should be calibrated.…”