Classification of pH levels using a mobile phone

Abstract

Mobile phones in today’s world are evolving into portable, powerful machine and turning from a luxury item into a necessity, equipped with easily accessible and downloadable applications. In this project, we employ the technologies of its camera functionality to come up with an automated application that allows the classification of pH on pH test strips on non-calibrated camera smartphones. The application takes a captured image of a pH test strip, dipped with any fluids, alongside with the color chart provided as input and predicts a pH reading of the test strip. This straightforward procedure makes it easy and generally convenient for the elderly as well as color blind users. The method used comprises edge techniques from image processing accompanied by clustering methods to locate the position of the color reference chart and the test strip. Thereafter, post color processing and filtering are done to the region of interest to evaluate the pH value of the test strip. The application is implemented in J2SE, and experiments cases conducted to test the robustness of the application, using a circular color chart instead of the rectangular chart, are shown to yields reliable results. With the proliferation of mobile camera phones in the market as a necessity, this solution is an accessible and convenient one, integrating pharmaceutical approaches into a portable application.