OFF-LINE HANDWRITTEN JAWI CHARACTER SEGMENTATION USING HISTOGRAM NORMALIZATION AND SLIDING WINDOW APPROACH FOR HARDWARE IMPLEMENTATION

The task of segmenting text into characters is a necessary preprocessing step in the development of most character recognition systems because incorrectly segmented characters are likely to be incorrectly recognized. The segmentation of off-line handwritten Jawi text poses a higher challenge due to its cursive nature and various writing styles. In this paper, histogram normalization and sliding windows are used for hardware implementation of real-time off-line handwritten Jawi script character segmentation. Existing algorithms for character segmentation are compared with the proposed method. The hardware design is presented along with justifications of the proposed approach. The main advantage of the proposed algorithm is its simple design which enables it to be implemented in hardware without requiring a large amount of resources. The character segmentation algorithm was implemented and the results show a 98% segmentation accuracy.