Brent Nelson
Papers
6
Total Citations
187
H-Index
6
About
Brent Nelson is a computer engineering researcher whose work sits at the intersection of reconfigurable computing, parallel architectures, and embedded vision systems. He is best known for his pioneering investigations into real-time optical flow computation, systematically benchmarking and comparing implementation strategies across FPGAs, GPUs, and massively parallel processor arrays. His landmark 2008 comparison study between FPGA and GPU architectures for optical flow has accumulated 72 citations, while his earlier 2007 FPGA-specific implementation work has drawn 56 citations — together establishing him as a leading voice on high-performance embedded image processing. Nelson's research addresses a fundamental challenge in robotic and autonomous systems: optical flow algorithms demand extraordinary computational throughput that conventional processors struggle to deliver in real time. By demonstrating viable FPGA-based solutions, he opened practical pathways for deploying these algorithms in time-critical applications. His work extends naturally into autonomous aerial platforms, including obstacle avoidance systems for unmanned air vehicles and embedded vision control for quadrotor helicopters — contributions that anticipated the modern UAV era. Across his body of work, Nelson has helped define the architectural trade-offs that engineers must navigate when deploying computationally intensive vision algorithms in embedded, power-constrained environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2FPGA-based Real-time Optical Flow Algorithm Design and Implementation56 citations · 2007
- 3
- 4
- 5An embedded vision system for an unmanned four-rotor helicopter10 citations · 2006
- 6Optical Flow on the Ambric Massively Parallel Processor Array (MPPA)9 citations · 2009