| When helicopter patrols transmission lines, the inspection images are the foundation for judging the working condition of transmission lines and the blur scale of image is also decisive factor for fault diagnosis and location of transmission lines, therefore, the restoration of blurred image is an essential part after patrolling.This paper presents a method for restoration of blurred image of transmission line based on aerial images. Firstly, judging the blur parameters of original blurred image (blur direction and blur scale), after the original blurred image being pre-processed, the blurred image will be restored by Lucy-Richardson, Blind deconvolution, Least squares and Wiener filter respectively. Secondly, via experimental results comparison, an improved Wiener filter method is proposed, Laplacian is applied to strengthen the edge information and contour feature of the image, then high frequency enhancement of Wavelet transform is employed to increase image depth further on the basis of Laplacian, the vector norm sum of image gray gradient G′k and Laplacian(8 neighborhood differential) operator sum Lk are introduced to evaluate the restored image quality, the two vary curves are painted in order to compare the restoration effect of original restored image and improved restored image . The test results show that the method can improve the image clarity efficiently. After recovering the blurred image, minimum mean absolute difference function (MAD), normalized correlation (NC) and invariant moments (IM) are used to match a spacer in the recovered image for comparing the three match methods, then choose invariant moments to match other six spacers in the recovered image and calibrate the spacers with Binocular Stereo Vision, set up the transmission line models, measure the sag and the distance between line and ground. At last, the trouble position of transmission line is located by tower number, Binocular Stereo Vision and GPS technology accurately. |
PDF Full Text Request