• Hrishikesh Vaidya
• Aditya Ajgaokar
• Shounak Daptardar
• Ankita Kadu

Text detection has often been a major field of research in computer vision. Previous approaches by the academics have achieved promising results with regard to scene text detection, but these systems usually lack in detecting text that is present in a document format. Traditional OCR engines generally fail when the text in the image doesn’t follow a linear format. The proposed system specializes in providing a fast, and accurate text detection system for images that contain systematic and spatial data which is challenging when processed through traditional OCR engines. Our methodical improvements result in efficient text detection through a system that is reliable, fast, and intuitive. It performs well on text that is non-linear in nature and is segmented throughout the image. In comparison to other text detection algorithms such as EAST text detection, our implementation easily outperforms the state-of-the-art text detection techniques. Our proposed system would be highly beneficial for businesses and corporations in handling their financial data.

text detection, invoice text detection, document OCR, openCV, image processing, deep learning

"A Novel Approach for Invoice Text Detection and Recognition", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.7, Issue 4, page no.467-469, April-2020, Available :http://www.jetir.org/papers/JETIR2004066.pdf

"A Novel Approach for Invoice Text Detection and Recognition", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.7, Issue 4, page no. pp467-469, April-2020, Available at : http://www.jetir.org/papers/JETIR2004066.pdf