Abstract The Javanese language has millions of speakers, making it one of the most widely-used languages in Indonesia and in the world at large. While the Javanese language still thrives in daily conversation, the number of users of Javanese script, the traditional script of the language, decreases every day. Many applications have been built to increase awareness and educate the public in the Javanese script, some of them using on stroke detection or image recognition to translate Javanese script for the user. $P stroke recognition was invented by Radu-Daniel Vatavu and friends in 2012 and has been tested to a high accuracy level in the detection of Latin alphabet characters and in sign detection. With the assumption that Javanese script can be recognized better, we evaluated the $P algorithm in detecting Javanese script using Cross-Validation and Micro- and Macro-average evaluation measure rule. A touchscreen Android-based application was built using Unity Engine.