Title :
Segmentation approach by learning: different image applications
Author :
Legal-Ayala, Horacio Andrés ; Facon, Jacques
Author_Institution :
PUC-PR, Pontificia Univ. Catolica do Parana, Curitiba, Brazil
Abstract :
We present a new segmentation approach by thresholding based on learning strategy. This strategy is based only on one image and its ideal thresholded version. A decision matrix is generated from each pixel and each gray level. At the moment of new image segmentation, the best solution for each pixel is evaluated by means of K nearest neighbors in the decision matrix. Comparative tests were performed on signature, fingerprint and magnetic resonance images.
Keywords :
biomedical MRI; decision theory; fingerprint identification; handwriting recognition; image segmentation; learning (artificial intelligence); medical image processing; statistical analysis; K nearest neighbors; decision matrix; fingerprint images; gray level; image segmentation; image thresholding; learning strategy; magnetic resonance images; signature images; Fingerprint recognition; Image analysis; Image segmentation; Lighting; Magnetic resonance; Nearest neighbor searches; Neural networks; Performance evaluation; Pixel; Testing;
Conference_Titel :
Image Analysis and Processing, 2003.Proceedings. 12th International Conference on
Print_ISBN :
0-7695-1948-2
DOI :
10.1109/ICIAP.2003.1234116