A smart phone-based robust correction algorithm for the colorimetric detection of Urinary Tract Infection.

This paper presents the preliminary work of developing a smart phone based application for colorimetric detection of Urinary Tract Infection. The purpose is to make a smart phone function as a practical point-of-care device for nurses or medical personnel without access to strip readers. The main challenge is the constancy of camera color perception across different illuminations and devices, which is the first step towards a practical solution without additional equipment. A reported black and white reference correction and a comprehensive color image normalization have been utilized in this work. Comprehensive color image normalization appears to be quite effective at correcting the difference in perceived color due to different illumination, and is therefore a candidate for inclusion in the further work.