Title :
The entropy strategy for shape recognition
Author_Institution :
Dept. of Math. & Stat., Massachusetts Univ., Amherst, MA, USA
Abstract :
We consider a computational strategy for shape recognition based on choosing “tests” one at a time in order to remove as much uncertainty as possible about the true hypothesis. The approach is compared with other recognition paradigms in computer vision and illustrated by attempting to classify handwritten digits and track roads from satellite images
Keywords :
entropy; image classification; image recognition; optical character recognition; remote sensing; computational strategy; computer vision; entropy strategy; handwritten digits; image classification; recognition paradigms; satellite images; shape recognition; track roads; Classification tree analysis; Decision trees; Entropy; Handwriting recognition; Matched filters; Performance evaluation; Roads; Satellites; Shape; Testing;
Conference_Titel :
Information Theory and Statistics, 1994. Proceedings., 1994 IEEE-IMS Workshop on
Conference_Location :
Alexandria, VA
Print_ISBN :
0-7803-2761-6
DOI :
10.1109/WITS.1994.513852
