This paper describes a new bottom-up method for document layout analysis. The algorithm was implemented in the CLIDE (Chemical Literature Data Extraction) system, but the method described here is suitable for a broader range of documents. It is based on Kruskal's algorithm and uses a special distance-metric between the components to construct the physical page structure. The method has all the major advantages of bottom-up systems: independence from different text spacing and independence from different block alignments. The algorithms computational complexity is reduced to linear by using heuristics and path-compression.

This paper presents a novel view of document processing, as being the reverse process to T/sub E/X. This concept simplifies the analysis of the physical structure of documents, and also suggests the use of a style file for layout recognition. An algorithm is given for both phases, layout analysis and layout recognition. The bottom-up layout analysis method employed is based on the Kruskal's algorithm and uses the distances between the components to construct the physical page structure. The algorithm is linear with respect to the number of the connected components. For layout recognition, a document style description language (DSDL) is introduced. This helps a fault-tolerant, recursive parsing algorithm to label the blocks of the document. The presented methods were designed to be used for scientific publications (papers, reports, books), but could be applied to a broader range of documents.