Fast and Low-Memory-Bandwidth Architecture of SIFT Descriptor Generation with Scalability on Speed and Accuracy for VGA Video

This paper describes an FPGA implementation which features a hardware-oriented Scale Invariant Feature Transform (SIFT) algorithm, a scalable architecture with high-speed mode and high-accuracy mode, and highly parallel datapath modules. The proposed FPGA implementation can generate a SIFT descriptor vector with 50 MHz for VGA resolution video (640 × 480 pixels) at 56 frames per second (fps). Our proposed implementation made the operating frequency and memory bandwidth a half or less than that of the conventional FPGA implementation and as a result, we achieved a system that can provide low power consumption.