Image detection method based on improved Canny algorithm

In the field of image edge detection, Canny algorithm has a better effect than other traditional edge detection algorithms in noise filtering and edge preserving. However, the traditional Canny algorithm has some defects in image preprocessing and adaptability. The Gaussian filter used in the traditional Canny algorithm is very poor in filtering out salt and pepper noise. The high and low threshold of the last step need to be set manually, which increases the time cost and lacks self-adaptability. This paper is proposed an improved Canny operator edge detection method, which uses adaptive median filter and guided filter instead of Gaussian filtering to smooth and denoise the image. Finally the improved Otsu algorithm is used to select the high and low thresholds to improve the overall adaptability. The simulation experiment results show that the improved Canny algorithm not only filters out most noises, especially salt and pepper noise, but also retains more image edge details, the identified image edges are smoother, and false edges are removed.