Text Detection and Character Recognition using Morphological Operation

Abstract

Character recognition from scanned images has been one of the dynamic and testing research zones in the field of picture preparing and design acknowledgment. It has various applications which incorporate, perusing help for visually impaired, bank checks and transformation of any written by hand report into basic content structure. In this venture an endeavor is made to perceive characters from English letter sets. and numbers by first segmenting them using bounding box and then recognizing each character individually. Each character data set contains 26 alphabets. For connected characters we have utilized morphological operation that method pictures subject to shapes. Morphological operations apply an arranging segment to an info picture, making an image of a comparative size. For the segmentation of cursive characters watershed algorithm can be used. For the purpose of character recognition template matching technique is used. The outcomes demonstrate that the proposed framework yields great acknowledgment rates which are practically identical to other comparable plans for character division and acknowledgment.