In this thesis, we present a real-time feature detection and object tracking system that is implemented on a low cost FPGA architecture. The algorithms are simulated in various environments to verify the real-time performance. A high-level programming language, Handel-C, is used to design and simulate in a DK1 design environment, and the design is synthesized by Xilinx WebPack synthesis tool. The final design is implemented on a low cost Xilinx Spartan-II-based FPGA device. The thesis also tries to implement the various algorithms in modular ways and tests these configurations on the FPGA device separately. The final design is implemented by merging all modular code in real-time.