Abstract
Despite the maturity of optical character recognition (OCR) technology, reading data from screens such as 7 segments and LCD screens possess some challenges such as identification, position and orientation for conventional OCR to be used directly for recognizing values from screens belongs to legacy manufacturing machines. Very often, these legacy machines do not have programmable interfaces and network connectivity thus the only way to capture its data is through optical means. This work, introduces the use of QR code (Quick Respond Code) labels as marker to guide the pre-processing steps of the capture image before using conventional OCR such as Tesseract to recognize the value in the image. From the QR code image, the system can detect if the capture image is tilted or slanted. This work shows that with such information the image can be processed and attain higher accuracy.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
