Low-cost, high-speed computer vision using NVIDIA's CUDA architecture.
Seung In ParkSean P. PonceJing HuangYong CaoFrancis K. H. QuekPublished in: AIPR (2008)
Keyphrases
- low cost
- high speed
- computer vision
- low power
- parallel implementation
- real time
- gpu implementation
- object detection
- object recognition
- compute unified device architecture
- management system
- computer graphics
- pattern recognition
- general purpose
- optical flow
- vision system
- graphics hardware
- structure from motion
- graphics processors
- image understanding
- times faster
- image features
- frame rate
- shared memory
- computer vision algorithms
- data acquisition
- single chip
- high speed networks
- image sequences
- reconfigurable hardware
- machine learning