Application of Improved Canny Algorithm in Image Edge Detection of Transmission Lines

Aiming at the incoherence and inaccuracy of traditional canny algorithm for transmission line image edge detection, an improved Canny algorithm is proposed for transmission line image edge detection. Firstly, the adaptive median filter is used instead of the traditional Gaussian filter to denoise the image, and the edge details can be preserved better. To solve the problem of single template of traditional Canny operator, gradient templates in horizontal, vertical, 45° and 135° are used to calculate the image gradient, which reduces the error detection rate and edge loss. Then the non-maximum value is suppressed by linear interpolation to make the edge information more accurate. Finally, in view of the limitation of traditional Canny operators artificially setting high and low thresholds, the maximum interclass difference method is used to calculate the high and low thresholds according to the image gray scale, so as to enhance the adaptability. The experimental results show that in comparison with traditional Canny algorithm, the improved Canny operator has better continuity and integrity in image edge extraction of transmission lines, and has obvious advantages.