Acoustic fingerprinting is a modern technique used for audio recognition, where short excerpts of audio files are used as a compact "signature" for the underlying media content. In this thesis, we have focused on the design, development and implementation of such a service; an application for mobile devices, capable of querying a database with a large collection of acoustic fingerprints to identify multimedia content in real-time, mainly movies and TV-series.

Our finished framework is developed to be robust, and the methods employed are following industry standards. As of now, the finished library come in two flavors; one written for the cross-platform Mono framework, and one written for the Apple Foundation framework. To continue the development of the platform would surely create a good basis for a new research or bachelor project, in an interesting field with a lot of promise.

Research regarding real-time acoustic fingerprinting is limited, and redefining the scope of the study was done several times throughout the project. Scientific papers on near-real time acoustic fingerprinting, adapted to a real time system, make up the most of the literature used in this thesis, and we base most of our production work on these.

In the information gathering phase of the project, we selectively picked well-reputed papers from several sources, including ScienceDirect and IEEE. The information gathered proved reliable, and have been used in several near-real time acoustic fingerprinting systems previously.