Irwansyah, Arif: Heterogeneous computing systems for vision-based multi-robot tracking. 2017
Inhalt
- Introduction
- Vision-based Robot Tracking Computing System
- Basic Concept of Vision-based Robot Tracking System
- Related Work
- CPU-based Computing System
- FPGA Accelerated Computing System
- GPU Accelerated Computing System
- FPGA-GPU Accelerated Computing System
- Hardware Accelerators in Vision Processing
- Summary
- Vision-based Multi-Robot Tracking with Heterogeneous Computing Systems
- Heterogeneous Computing System
- Architecture and Design Flow
- Vision-based Multi-Robot Tracking Algorithm
- Summary
- Implementation in FPGA-accelerated Heterogeneous Computing System
- FPGA-CPU Hardware Environment Description
- Algorithm Implementation
- Vision Processing Module Implementation in FPGAs
- Multi-Camera GigE Vision Frame Grabber Module
- Object Segmentation
- Edge Filter Module
- Object Localization
- Resource Utilization
- Post Processing in Host PC
- Summary
- Implementation in GPU-accelerated Heterogeneous Computing System
- GPU-CPU Hardware Environment Description
- Algorithm Implementation
- CUDA Kernels Implementation
- Achieved Occupancy
- Post Processing in Host PC
- Summary
- Results and Analysis
- Conclusions and Outlook
- List of Figures
- List of Tables
- Abbreviations
- References
- Author's Publications