Research on Canny Edge Feature Detection Technology of Color Image Based on Vector Properties

Canny edge detection based on optimization algorithm has the advantages of high accuracy and high signal-to-noise ratio. However, because the Canny algorithm was based on gray image, some edge information will be lost during the process of converting color image to gray image. Aiming at this problem, the edge detection method of color image based on vector properties was proposed in this paper. According to the properties of vectors with direction and length, the image gradient calculation of Canny algorithm was improved from the angle and length of vector, in order to realize edge feature detection of color image directly. Compared with the results detected by the conventional Canny algorithm, this method can get clearer and more accurate edges. When combined with specific color markers in specific application scenarios, this method can obtain ideal detection results that only contains the edge of the marker.