Classification of off-line hand-written words into upper and lower cases

The paper presents an efficient technique for classification of offline handwritten words into upper and lower case using principal components (PC). The technique consists of two phases. For each word, in feature extraction phase, first the boundary points of the word are extracted, then twenty-six features including global, local, region and dominance features are extracted using the contour information. In the classification phase, a discriminate function based on the PC adapted by our system, is introduced to integrate the extracted features and classify words into upper and lower case. Experimental results show that the system achieves an 83% correct word case classification for about 2240 test words randomly selected from a 3226 data set obtained from 12 writers