Key points are not available for this paper at this time.
Edge detection is a critical component of many vision systems, including object detectors and image segmentation algorithms. Threshold setting is crucial for edge detection. Traditional threshold setting method did not use the spatial information of edge, and the detected edges are either missing or have a large number of false edges. Based on the analysis of the change trend of edge connectivity and the number of edge segment as the threshold decrease, we firstly propose an adaptive threshold setting method according to the image self-information. Then, using the adaptive threshold, the gradient map created by first order differential operator and non-max suppression is binarized into edge map. Compared with other traditional single-threshold based edge detection methods, the proposed method can detect edges that are more complete and fewer fake edges, and this method is more stable and reliable.
Mo et al. (Mon,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: