Title :
Supervised texture segmentation through a multi-level pixel-based classifier based on specifically designed filters
Author :
Melendez, Jaime ; Girones, Xavier ; Puig, Domenec
Abstract :
This paper presents a new, efficient technique for supervised texture segmentation based on a set of specifically designed filters and a multi-level pixel-based classifier. Filter design is carried out by means of a neural network, which is trained to maximize the filters´ discrimination power among the texture classes under consideration. Texture features obtained with these filters are then processed by a classification scheme that utilizes multiple evaluation window sizes following a top-down approach, which iteratively refines the resulting segmentation. The proposed technique is compared to previous supervised texture segmenters by using both synthetic compositions and real outdoor textured images.
Keywords :
filtering theory; image classification; image segmentation; image texture; multi-level pixel-based classifier; real outdoor textured images; specifically designed filters; supervised texture segmentation; synthetic compositions; Adaptive filters; Conferences; Feature extraction; Filter banks; Gabor filters; Image segmentation; Support vector machines; Specific texture filters; Supervised texture segmentation; multi-level classification; neural networks;
Conference_Titel :
Image Processing (ICIP), 2011 18th IEEE International Conference on
Conference_Location :
Brussels
Print_ISBN :
978-1-4577-1304-0
Electronic_ISBN :
1522-4880
DOI :
10.1109/ICIP.2011.6116147