A method in applying OCR in processing of thermal images
W.T. Chan, T. Y. Lo, Chih Ping Tso, Kok Swee Sim · 2011
The processing of thermal images that lack temperature scale information as metadata can be performed with the aid of optical character recognition. Optical character recognition scripting can automate the determination of the temperature of locations in the image from their color information without the need for manual input of the temperature range. This paper aims to produce a working program that can perform this, by using methods that compare characters captured from the thermal image with a library of identified characters and assigning association between them, and then contrast this program with one that requires manual input. The former program is found to produce greatly reduce the time needed to process thermal images by replacing the manual input stage with optical character recognition scripts.