A Cloud Based Object Recognition Platform for IOS

Food recognition is not only a challenging task but also important for patients and travelers. In order to provide a good recognition capability, we propose a cloud based object recognition platform for IOS platform, in which an improved 2DPCA algorithm is used for object recognition. A Hadoop based cloud server is built for this platform. To show the effectiveness of this platform, a series of tests are finished using ALOI (Amsterdam Library of Object Images) database. These tests show that our approach has a better performance in multi-radians and multi-illumination environment compared with existing approaches.