Image edge analysis and application is a classic problem in image analysis and an important preprocessing method in image follow-up processing. In order to use subpixel edge extraction method to extract edge pixels and establish automatic curve segmentation and fitting model, this paper first smooths the image through Gaussian filter and sharpened the image to highlight the contour features. Then subpixelation is carried out, the optimized inverse Gaussian edge test operator is established, and the edge contour curve is obtained. Then the curvature of each point is calculated by the gradient algorithm of the discrete contour points, and the module segmentation is completed by the input of the module segmentation state machine, the curve parameters are determined, and the image contour is vectorized. In the complex environment, this research carries out contour recognition and vector information extraction on bitmaps taken by cameras, which has important reference value in the field of industrial machine vision to improve the accuracy of photo contour recognition and reduce the space occupied by pictures. Moreover, the segmentation fitting of the second step can provide clues for further detection of whether there is a bump in industrial parts.
Image contour edge segmentation and fitting based on Anny-Gaussian edge test operator
2023-10-11
2786843 byte
Conference paper
Electronic Resource
English
Generalizing edge detection to contour detection for image segmentation
British Library Online Contents | 2010
|"Nieszczęsny (auto)portret" Anny Bilińskiej-Bohdanowicz
DataCite | 2012
|A Bayesian Multiple-Hypothesis Approach to Edge Grouping and Contour Segmentation
British Library Online Contents | 1993
|Edge affinity for pose-contour matching
British Library Online Contents | 2006
|Edge detection by curve fitting
British Library Online Contents | 1995
|