Document segmentation and classification with top-down approach

A robust algorithm for document segmentation and classification has been implemented. In this algorithm, a top-down approach is adopted due to its fast processing speed and reliability. The approach accounts for every portion of the processed page. It is also capable of image display at different scales and allows editing of classified blocks such as the changing, merging and splitting of blocks. The segmentation and classification processes are explained using examples. A large number of sample document images have been tested with satisfactory results.