Word Spotting from Handwritten Kannada and Roman Text Documents
Abstract: In this paper, we reported the efficient method for word spotting from handwritten Kannada and Roman text documents. Our main aim is to classify the given word whether it?s Kannada or English based on keyword. Firstly, words are extracted from scanned document images. Then Radon transform with different orientation and texture features such as local entropy, local range, and local standard deviation are used as features from each word image. Popular Euclidean distance is used for classification of the document and obtained average 99.25% accuracy in word spotting form document images.
Keywords: Handwritten Script, Kannada, Roman, Radon Transform, Cosine, OCR
How to Cite?: Veershetty C, "Word Spotting from Handwritten Kannada and Roman Text Documents", Volume 12 Issue 8, August 2023, International Journal of Science and Research (IJSR), Pages: 217-220, https://www.ijsr.net/getabstract.php?paperid=SR23730131126, DOI: https://dx.doi.org/10.21275/SR23730131126