Unsupervised Color Texture Feature Extraction and Selection for Soccer Image Segmentation

In this paper, we describe a new approach for color texture featureextraction and selection. We definecolortexturefeatures as texture features which are computed by taking into account the color components of the pixels. We determine the most discriminatingcolor texture featuresamong a multidimensional set of color texture features by means of an iterative feature selection procedure associated to an information criterion. This procedure analyses images which are classified by a competitivelearning scheme. Soccer image segmentationis achievedby pixel classification. The classification algorithm takes intoaccount these color texturefeatures which areprocessed in the neighborhoodof the pixels. We apply our new unsupervised approach to soccer images segmentation.

• Type: Conference
• Year: 2000
• Where: ICIP
• Authors: Nicolas Vandenbroucke, Ludovic Macaire, Jack-Gérard Postaire
• Keywords: Color Texture Featureextraction, Color Texture Features, Discriminatingcolor Texture, ICIP 2000, Image Processing