A pipeline hog feature extraction for real-time pedestrian detection on FPGA.
Vinh NgoArnau CasadevallMarc CodinaDavid Castells-RufasJordi CarrabinaPublished in: EWDTS (2017)
Keyphrases
- pedestrian detection
- real time
- feature extraction
- human body
- object detection
- histograms of oriented gradients
- human detection
- low cost
- crowded scenes
- benchmark datasets
- advanced driver assistance systems
- detection rate
- occlusion handling
- face recognition
- parallel architecture
- fpga device
- computer vision
- face detection
- image classification
- feature vectors
- image processing
- texture classification
- viewpoint
- computational complexity
- pedestrian tracking