Canny Threshold Selection Algorithm Based on the Second Derivative of Image Gradient

Image edge contains abundant information crucial for many visual systems, such as target detection, image segmentation and etc. Traditional Canny operator exclusively applicable to gray images, and it cannot use color information effectively in multi-color images. Nevertheless, traditional Canny operators require setting the high and low threshold manually, disabling adaptive edge detection and bringing other problems, such as amplification of background edge. In this paper, an adaptive edge detection algorithm based on Canny (AEDAC) in multi-color images is proposed. Firstly, the AEDAC use first-order statistical properties of the histogram of images to select parameters of the Gaussian filter adaptively, effectively removing the noise and reducing the influence of unreasonable parameter setting on edge detection. Secondly, the threshold selection method based on the two derivative of image gradient is adopted to adaptively select appropriate threshold according to characteristics of images. Experimental results show that the AEDAC weakens defects of traditional Canny operator and extracts edge of multi-color images effectively.