To search, Click below search items.


All Published Papers Search Service


An Approach to Recognize Bangla Digits from Digital Image


Abdul Kadar Muhammad Masum, Mohammad Shahjalal, Iqbal Hasan Sarker, Faisal Faruque


Vol. 11  No. 6  pp. 248-252


Pattern recognition is important in converting digital document image into electronic document. Today there are some OCR softwares to read the documents from scanned papers. We have designed a process to recognize the Bangla digits from digital images. Bangla digits are discrete when they are used in documents to write numbers. Each character has its own shape and design. We have used these characteristics of Bangla digits to recognize them from digital images. We have applied different morphological operation and segmentation techniques to separate each digit or character from the image. Then we have just drawn a single bit outline of the character, and generated a sequence based on the traversing path of the border outline. Then we applied a longest common subsequence operation between the generated sequence of the character and the pre-defined sequence of the 10 digits and detected the unknown Bangla digit. This is a sequence-based method.


Digital image, Image processing, Bangla digit, LCS(Longest Common Subsequence), DFS(Depth First Search)