A modified recursive x-y cut algorithm for solving block ordering problems
Phaisarn Sutheebanjard, Wichian Premchaiswadi · 2010
To achieve the best results from an OCR system, the pre-processing steps must be performed with a high degree of accuracy and reliability. There are two critically important steps in the OCR pre-processing phase. First, blocks must be extracted from each page of the scanned document. Secondly, all blocks resulting from the first step must be arranged in the correct order. One of the most notable techniques for block ordering in the second step is the recursive x-y cut (RXYC) algorithm. This technique works accurately only when applied to documents with a simple page layout but it causes incorrect block ordering when applied to documents with complex page layouts. This paper proposes a modified recursive x-y cut algorithm for solving block ordering problems for documents with complex page layouts. This proposed algorithm can solve problems such as (1) the overlapping block problem; (2) the blocks overlay problem, and (3) the L-Shaped block problem.