Abstract
While the number of uses for ID(Identification) OCR(Optical Character Recognition) is increasing and the scope of its use is increasing, it has limitations in increasing the probability due to the problem of not being able to controlexternal conditions. The feedback currently being provided is not scientific, so it is unable to provide appropriate guidance to users. This study created a formula to avoid veiling reflection, which is the main cause of poor OCR, and allowed the angle of lighting to be adjusted according to the distance between the ID card and the camera. In addition, the BRDF(Bidirectional Reflectance Distribution Function) of the ID card was measured and simulated, and the point where the color luminance contrast was greatest was presented at an angle of 36 degrees between the light and the ID card, and this was verified in an actual situation. Through this, points where OCR can be good in an artificial lighting environment were presented in a scientific way, and machine learning was additionally conducted to provide a way for users to conveniently find places with a high OCR probability.
Sign-in/Register to access full text options 
Free) 
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
