FPGA-Based Real-Time Image Processing Accelerator for Edge IoT Applications

Objective: To design, implement, and validate a low-latency FPGA accelerator for real-time image processing on resource-constrained edge devices.

Real-world problem: Many industrial, medical, and smart city applications in India require real-time image processing at the edge—for example, in traffic surveillance, smart agriculture, and biomedical monitoring—where cloud connectivity is limited or latency is critical. Traditional CPUs or MCUs on edge devices struggle with the computational demands of image processing, resulting in slow response times and high power consumption.

Proposed solution: This project aims to develop a field-programmable gate array (FPGA)-based hardware accelerator tailored for real-time image processing tasks such as object detection, edge detection, and filtering. The system will be designed as a plug-in module for existing edge IoT platforms, leveraging parallel hardware logic to dramatically improve processing speed and energy efficiency over traditional software approaches.

Key features and deliverables: The project will deliver a working prototype comprising a low-cost FPGA development board (such as Xilinx Artix-7), an interfaced camera module (e.g., OV7670), and a software stack for integration with edge devices (such as Raspberry Pi or ARM-based SoCs). Demonstrations will include live video feed processing with measurable improvements in latency and power usage, along with detailed documentation, cost analysis, and industry-aligned paper/report.

Industry impact: The accelerator prototype can be adapted for Indian industry use cases in surveillance, agriculture, and healthcare, enabling affordable, scalable, and energy-efficient edge vision solutions. The design is modular, standards-compliant (VHDL/Verilog), and suitable for commercial prototyping or further academic research.