This document describes a mobile application that uses handwriting recognition to reduce language barriers for tourists. The application allows tourists to take photos of text in Arabic and receive translations in their own language. The system is based on Hidden Markov Models (HMMs) and was trained on a database of handwritten Tunisian village names. When tested on printed Lebanese village names and restaurant menus, the system achieved an error rate of only 2% without any printed text in its training data. Future work will further study the simultaneous recognition of handwritten and printed text and explore adding printed texts to the training set.