Title :
Multiscale segmentation of unstructured document pages using soft decision integration
Author :
Etemad, Kamran ; Doermann, David ; Chellappa, Rama
Author_Institution :
Center for Autom. Res., Maryland Univ., College Park, MD, USA
fDate :
1/1/1997 12:00:00 AM
Abstract :
We present an algorithm for layout-independent document page segmentation based on document texture using multiscale feature vectors and fuzzy local decision information. Multiscale feature vectors are classified locally using a neural network to allow soft/fuzzy multi-class membership assignments. Segmentation is performed by integrating soft local decision vectors to reduce their “ambiguities”
Keywords :
document image processing; feature extraction; feedforward neural nets; fuzzy set theory; image segmentation; image texture; uncertainty handling; wavelet transforms; context dependent classification; document processing; document texture; feedforward neural network; fuzzy membership assignments; multiscale feature vectors; multiscale segmentation; soft decision integration; unstructured document pages; wavelet packets; Background noise; Context; Electronic mail; Fuzzy neural networks; Graphics; Image segmentation; Image storage; Information analysis; Layout; Neural networks; Performance analysis; Wavelet analysis; Wavelet packets; White spaces;
Journal_Title :
Pattern Analysis and Machine Intelligence, IEEE Transactions on
