�ݺ�ߣ

Developing Video Signal
Processing Algorithms for
Embedded Vision
Systems
Shogo MURAMATSU
Dept. of Elec. & Electronic Eng.
Niigata University

Contents
? Introduction to Embedded Vision
? Tutorial Material with MATLAB?/Simulink ?
?EmbVision Tutorial
? Example Activities
?Smart Dormitory Program (PBL Example)
?NSOLT Project (Research Example)
? Conclusions
2015/8/26 SISA 2015 @ Chiba Institute of Tech.
? Tutorial Material with MATLAB?/Simulink?
?EmbVision Tutorial
? Conclusions

The IoT Era is Beginning.
? Diversification of Sensing Environment
Embedded Systems will be Deployed Broadly.

Seeing is Believing.
? Rapid Proliferation of Vision Systems
Embedded
System
Computer
Vision
Embedded
Vision
Embedded Vision Systems gather Attention.

Contents
? Tutorial Material with MATLAB ? /Simulink ?
? EmbVision Tutorial
? Smart Dormitory Program (PBL Example)
? NSOLT Project (Research Example)
? Conclusions

From DSP First to SI First
? DSP First: A Multimedia Approach (1998)
? Presents basic DSP concepts in an intuitive
style by multimedia signals with MATLAB.
? Evolution of MATLAB makes the System
Integration (SI) First Approach available.
? Embedded System Development
? Web + Database Management
? Network Connection
DB

EmbVision Tutorial
? Educational Material
for new members in
our Lab.
(Junior or Senior)
? Estimated to take 12H
SISA 2015 @ Chiba Institute of Tech.2015/8/26
Today, condensed to
50min!
http://msiplab.eng.niigata-u.ac.jp/embvision/en/

Demo with Raspberry Pi?
? Gradient Filtering on
Raspberry Pi
(Final Exercise)

Overview of EmbVision Tutorial
? One can implement user-defined System
objects on Raspberry Pi.
Part 1 Part 2 Part 3 Part 4 Part 5 Part 6
Intensity
Adjustment
Color space
Conv.
Grad.
Filter
System
Object?
Video
Stream
Process.
MATLAB
System
block
Unit
Test
Raspberry
Pi
MATLAB Simulink

Part 1:
Image I/O and Pixel Processing
? Target
? Read, display, and write images
? Simple pixel processing
? Exercises
? Intensity Adjustment
? Color Space Conversion
RGB2
HSV
HSV2
RGB
? �� 2?

Part 2:
Filtering and Frequency Analysis
? Target
? 1-D signal and 2-D image filtering
? Frequency analysis
? Exercises
? Horizontal Differential Filter
? Magnitude and Direction of Gradient

Part 3:
Class Definition and Unit Testing
? Target
? Object-oriented programming in MATLAB
? Unit testing framework in MATLAB.
? Exercises
? HSV2RGB Class
? Gradient Filter Class

Part 4: Video Stream Processing
- MATLAB -
? Target
?Read, display and write videos in MATLAB
?Video stream processing in MATLAB
? Exercises
?Sobel Gradient Filter
?Frame Difference

- Simulink -
? Target
? Read and display videos with Simulink
? Use MATLAB System blocks on Simulink
? Video stream processing with Simulink
? Exercises
? Prewitt Gradient Filter
? Sobel Gradient Filter

- Raspberry Pi? -
? Target
? Simulate Simulink model in external mode
? Deploy Simulink model on Raspberry Pi
? Exercises
? Sobel Gradient Filter
? Deploy a created new model
(Option)

DIY
? List of Expected Extension Examples
? Raspberry Pi Camera Module Control
? Fixed-point Implementation
? Parallel Implementation (for Quad Core on RasPi2)
? Detection & Recognition of Objects
? Speech & Audio Signal Processing
? GPIO/I2C Controls
? Network Application Development
? etc.

Contents
? Tutorial Material with MATLAB?/Simulink?
?EmbVision Tutorial
? Conclusions

Smart Dormitory Program
? Project-Based Learning (PBL) Program at Facul.
of Eng., Niigata Univ.
? Financially Supported by MEXT
? MEXT: Ministry of Education, Culture, Sports, Science and
Technology
? 15 Research Groups are working on their research
activities.
? The 2nd group is Cyber Physical System (CPS)
Dormitory
? Members are from different grades and departments.
? Develop an IoT App. with embedded vision systems

CPS Dormitory
? Develops a System for Estimation and Visualization of
Customer Congestion at a Student Cafeteria
? BeagleBone Black boards as Sensor Nodes
? ThingSpeak as an IoT data collection plathome
Installation of a Sensor Node Deployment Map of Sensor Nodes

Presentation at the 4th MEXT
Science Inter-College
@Kobe Convention Center
29th Feb. - 1st Mar., 2015

? Nonseparable Oversampled Lapped Transform
? Sparsity-Aware Image and Volume Data Restoration
(a)
(b)
NSOLT Project
[Muramatsu,ICASSP2014]
Simulink Model

Zynq? Implementation
? Xilinx?��s All Programmable SoC
?ARM? CPU and FPGA are connected to each
other through AXI4 interface on a single chip
Xilinx��s Zynq Inside of Zynq

Workflow of Co-implementation
Generation of HDL
(HDLCoder?)
Generation of C
(Embedded Coder? )
Configuration Build
Zynq
2015/8/26
System Model of NSOLT
(MATLAB/Simulink)
Isolation of HW and SW in the model
SISA 2015 @ Chiba Institute of Tech.

Conclusions
? This tutorial explained
? How to develop and evaluate video signal
processing algorithms with MATLAB/Simulink
? How to implement visual applications on
embedded systems with MATLAB/Simulink
? Two example activities were introduced
? CPS Dormitory Prog. as an Educational Activity
? NSOLT Project as a Research Activity

�ݺ�ߣ

Developing Video Signal Processing Algorithms for Embedded Vision Systems

More Related Content

Developing Video Signal Processing Algorithms for Embedded Vision Systems