Image edge detection method based on anisotropic diffusion and total variation models

Autor: Ali Abdullah Yahya, Jieqing Tan, Benyu Su, Kui Liu, Ali Naser Hadi
Jazyk: angličtina
Rok vydání: 2019
Předmět:
Zdroj: The Journal of Engineering (2019)
Druh dokumentu: article
ISSN: 2051-3305
DOI: 10.1049/joe.2018.5345
Popis: In this study, a novel image edge detection technique based on the combination of total variation (TV) and anisotropic diffusion (PM) models is presented. In the proposed technique, the authors first use the gradient magnitude to eliminate the noise, then utilise the adaptive weight function to detect the edges of the image. The adaptive weight function has a high ability to adapt and change according to the areas information (edges or flats areas). More specifically, TV filter is applied on the areas which suffer from double and false edges, whereas, anisotropic diffusion filter is applied on the areas which suffer from weak and discontinuous edges. Applying TV filter on the double edges areas will allow one to remove most of the false edges, and thus to obtain much sharper edges. While, applying anisotropic diffusion filter on the discontinuous edges areas will lead to obtaining robust and continuous edges. Consequently, less false edges besides high localisation accuracy were obtained. Experimental results demonstrate the superiority of the new approach in terms of removing the false edges and improving the localisation accuracy of the edges. As objective quantitative performance measures, the peak signal-to-noise ratio (PSNR) and Pratt's figure of merit (FOM) were used.
Databáze: Directory of Open Access Journals
Externí odkaz: