Design of a Smart Embedded Vision System Based on FPGA for Medical Applications

Abstract

Applications in fields like medical, robots, and automobiles must have embedded vision with the optimized design metrics in today’s smart environment. A tiny computer and camera can be used to create a tiny vision system for the specific purpose; this kind of system is called an embedded vision system. For better design metrics, this project involves developing an embedded vision system and putting it to use in real-time applications. Design criteria that are optimized include low price, rapid development, compact size, and swift speed. The proposed embedded vision system is implemented with the ZYNQSOC. The ZYNQ SOC combines the software programmability of a CPU with the hardware programming of such an FPGA to give high adaptability, excellent performance, scalability, and low power. The embedded vision system was tested by attaching the OV7670 camera to the zedboard. The Zedboard’s client-server architecture enables real-time continuous data streaming to the Computer via Ethernet. The hardware platform for the suggested embedded vision system is developed with built-in IP cores in VHDL in Vivado, C-language application software written in SDK, and Petalinux on Ubuntu. This platform is required for applications based on embedded vision systems. Applications for the proposed biomedical clever integrated vision system are available.