Title :
Corrected unwrapping method based on the tangential and radial distortion centre for the panoramic annular image
Author :
Xinzheng Wang ; Xiongzhu Bu ; Bo Yang ; Guijuan Li ; Jing Yu
Author_Institution :
Sch. of Mech. Eng., Nanjing Univ. of Sci. & Technol., Nanjing, China
Abstract :
A novel corrected unwrapping method is proposed for panoramic annular images based on the catadioptric panoramic imaging system. The method is comprised of four key steps. Firstly, the distortion centre parameters on the panoramic annular image are determined, including the tangential distortion centre by using the least-square fitting (LSF) method and the radial distortion centre by extracting feature points. Secondly, the affine transformation relation between panoramic annular image and cylindrical panoramic image is constructed based on the centre parameters and the geometric properties of panoramic image. Then the annular image pixels are replicated into the cylindrical panoramic image by means of the bilinear interpolation. Finally, the frame aspect ratio for the final image unwrapping is corrected. The motivation for developing this technique is to solve the distortion problem from the centre deviation, and ultimately to promotes its adoption in pipe non-destructive testing systems. Experimental results have proved the feasibility, validity and higher accuracy of the unwrapping method.
Keywords :
affine transforms; distortion; feature extraction; interpolation; least squares approximations; LSF method; afflne transformation relation; annular image pixels; bilinear interpolation; catadioptric panoramic imaging system; centre deviation; corrected image unwrapping method; cylindrical panoramic image; feature point extraction; frame aspect ratio; geometric property; least-square fitting method; panoramic annular image; pipe nondestructive testing systems; radial distortion centre parameter; tangential distortion centre parameter;
Journal_Title :
Image Processing, IET
DOI :
10.1049/iet-ipr.2014.0301