Automatic separation of text from background in scanned images of complex documents

摘要: A system that converts a scanned image of complex document into an where text has been preserved and separated from the background. The first subdivides blocks then examines each block pixel by to construct histogram gray scale values pixels. is partitioned first, middle last regions. If one or more peaks occur in regions, single peak occurs within region, pixels are reexamined determine frequency occurrence having level nearby which have region peak. this high, assumed be background information. After determining threshold, rescans applying threshold separate information block.