摘要: Document image processing has become an increas- ingly important technology in the automation of office documentation tasks. Automatic document scanners such as text readers and op- tical character recognition (OCR) systems are essential compo- nent capable those One problems this field is that to be read not always placed correctly on a flat-bed scanner. This means may skewed scanner bed, resulting image. skew detrimental effect analysis, under- standing, segmentation recognition. Conse- quently, detecting correcting it issues realizing practical reader. We describe new algorithm for detection. then compare performance results detection other published methods from O'Gorman, Hinds, Le, Baird, Postl, Ak- iyama. Finally, we discuss theory different approaches taken solve problem docu- ments. The correction propose been shown extremely fast, with run times averaging under 0.25 CPU sec- onds calculate angle DEC 5000/20 workstation. © 1996 SPIE IS&T.