A color image segmentation method based on seeded region growing and visual attention

• Autorzy: Ouerhani N. , Archip N. , Huegli H. , Erard P.
• Konferencja: International Conference on Computer Analysis of Images and Patterns (9 ; 2001 ; Warszawa)
• Języki publikacji: EN

Abstrakty

EN

This paper reports a color image segmentation method based on a seeded region growing technique (SRG) and guided by a saliency-based visual attention algorithm. Inspired by biological vision the purely data-driven model of visual attention is built around the feature, conspicuity and saliency maps. Using chromatic as well as unchromatic scene features, it, automatically, generates a set of regions-of-interest (ROIs), which represent the most visually-salient locations of the image. The automatically selected points are then used as seeds by the region growing algorithm to segment the conspicuous parts of the scene, using a color homogeneity criterion. A snakes-based technique is then used to improve the contours of the segmented regions. The results reported in this paper clearly shoe the effectiveness of the considered model of visual attention to detect the salient locations in color images.

Słowa kluczowe

EN

image color analysis; image segmentation; visual perception

PL

analiza koloru obrazu; percepcja wizualna; segmentacja obrazu

• Wydawca: Instytut Telekomunikacji i Informatyki Uniwersytetu Technologiczno-Przyrodniczego w Bydgoszczy
• Czasopismo: Image Processing & Communications
• Rocznik: 2002
• Tom: Vol. 8, no 1
• Strony: 3--11
• Opis fizyczny: Bibliogr. 17 poz.,Rys., wykr.,

Twórcy

autor

Ouerhani N.

autor

Archip N.

autor

Huegli H.

autor

Erard P.