AR technology can be considered as mainly consisting of two aspects: identification of real-world object and display of computer-generated digital contents related the identified real-world object. The technical challenge of mobile AR is to identify the real-world object that mobile device's camera aim at. In this paper, we will present a location-based object identification algorithm that has been used to identify learning objects in the 5R adaptive location-based mobile learning setting. We will also provide some background of the algorithm, discuss issues in using the algorithm, and present the algorithm empowered mobile learning system and its implementation.