| 摘要 |
A method of managing storage in a document image database using document analysis to partition documents into logical regions and modified by reducing storage size of the regions using different reduction means according to various storage preference rules. Storage preference rules are intended to maintain high quality representations of important document information while reducing storage requirements at the expense of lesser important aspects of the document. In particular, the different reduction means applied to stored document images includes reducing sampling depth, reducing sampling resolution based on minimum font size, utilizing lossy and lossless compression schemes and discarding unimportant regions of document image. Over time, document analysis and modification can be repeated to further reduce the storage size of previously stored data files.
|
