Main Article Content
Abstract
This paper exhibits a total Optical Character Recognition (OCR) framework for camera caught picture/illustrations implanted printed archives for handheld gadgets. From the start, content locales are removed and slant adjusted. At that point, these areas are binarized what's more, fragmented into lines and characters. Characters are passed into the acknowledgment module. Trying different things with a lot of 350 analytical card pictures, caught by wireless camera, we have accomplished a most extreme acknowledgment exactness of 91.43%. Thought about to Tesseract, an open source work area based amazing OCR motor, present acknowledgment precision merits contributing. In addition, the created procedure is computationally productive what's more, expends low memory in order to be pertinent on handheld gadgets.