Face recognition using Local Binary Patterns Histograms (LBPH) on an FPGA - based System on Chip (SoC ) Nikolaos Stekas 1 , Dirk van den Heuvel 2 1. Delft University of Technology, Delft, Netherlands 2. TOPIC Embedded Products*, Best, Netherlands *[email protected]
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Face recognition using Local Binary Patterns Histograms (LBPH) on an FPGA-based
System on Chip (SoC)
Nikolaos Stekas 1, Dirk van den Heuvel 2
1. Delft University of Technology, Delft, Netherlands2. TOPIC Embedded Products*, Best, Netherlands
Face Recognition
ID: 5235
• Evolution and Importance
• Challenges
Real Time Execution
High Recognition Accuracy
…LBPInput Image
Regions HIGH ACCURACY
SMALL DATABASE
SIMPLE COMPUTATIONS
Input Face Image Feature ExtractionDatabase Features
ComparisonRecognition
• LBPH
• Manhattan Distance
Δ = | x - y |
Implementation• Platform
• System Overview
Implementation• Platform
• System Overview
FPGA ACCELERATION
ARM SIMPLICITY
Results
0
2
4
6
8
10
12
14
16
18
LBPH MD Overall Real-Time
Exec
uti
on
Tim
e (m
s)
2.602
17
25.9
0
5
10
15
20
25
30
Current Work CUDA OpenCL
Exec
utu
ion
Tim
e (m
s)
1. Execution Time
2. Accuracy 3. Utilization
Thank you!
http://topicembeddedproducts.com/
Related Documents
