Distributed 3D Object Recognition System Using Smartphones

Object recognition and scene classification are generally considered one of the most important challenges in computer vision community, where, object recognition is a process of finding and identifying objects in a digital image or video sequence. One of the main problems in recognizing 3D object is extracting stable and consistent features vectors under different conditions, such as camera viewpoint, illumination and cluttered background. In addition, Processing and memory capacity of Smartphones still restrict the computational capacity of object recognition programs. In this paper, we propose a distributed 3D object recognition system to overcome computational capacity problem and improve scalability of objects that will simply be recognizable. The paper also proposes the use of k-Nearest Neighbors classifier with Speeded Up Robust Features algorithm to solve the problem of extracting stable and consistent features vectors. The system is remarkably capable of adapting to different network configurations and the wireless bandwidth, and improving the performance of recognizing multiple 3D objects using Smartphones devices. Keywords—Scale Invariant Feature Transform; Speeded Up Robust Features; k-Nearest Neighbors.