TRIPOD: Computer Vision for Classroom Instruction and Robot - - PowerPoint PPT Presentation

▶
tripod computer vision for classroom instruction and
SMART_READER_LITE
LIVE PREVIEW

TRIPOD: Computer Vision for Classroom Instruction and Robot - - PowerPoint PPT Presentation

Jan 10, 2023 461 likes •625 views

TRIPOD: Computer Vision for Classroom Instruction and Robot Construction Paul Y. Oh Drexel University Mechanical Engineering Program for Robotics, Intelligent Sensing and Mechatronics Presentation AAAI Spring Symposium March 22-24, 2004 1

slide-1
SLIDE 1

1

Paul Y. Oh

Drexel University – Mechanical Engineering Program for Robotics, Intelligent Sensing and Mechatronics

TRIPOD: Computer Vision for Classroom Instruction and Robot Construction

Presentation AAAI Spring Symposium March 22-24, 2004

slide-2
SLIDE 2

Motivation: Research in Seeing Robots and Mechanisms

Visually Servoed Tracking IROS 2002

Head-Mounts for Alleviating Motion Sickness & Balance Disorders

LEAP: Low Elevation Aerial Photography ICRA 2003 CQAR: Closed Quarter Aerial Robots IROS 2003 Visually Servoed Bipeds

slide-3
SLIDE 3

Motivation: Undergraduate Robotics and Outreach

US FIRST – High Schools and ASME Future Drexel Teams?

slide-4
SLIDE 4

Computer Vision for the Classroom?

Must learn Matlab Robust VFW Windows Matlab I/A Toolbox No video Robust VFW Unix, Windows Matlab I/P Toolbox Proprietary Turnkey Proprietary Windows Coreco Sherlock Expensive Documentation Proprietary Windows Matrox MIL Setup difficult Open, e-Community VFW Windows Intel Open CV Setup difficult Open, Libraries Various Unix X-Vision, X-Vision 2 Cons Pros Hardware Platform Vision Package

Common Denominators:

  • ANSI C/C++
  • Access to Windows PC
  • No DirectX, ActiveX, MFC
  • Cost constraints
  • Affordable
  • Robust
  • Customizable

Design Parameters

  • Teach Computer Vision!
  • Microsoft always changing!
slide-5
SLIDE 5

Computer Vision for Robot Construction?

  • Mechanism Dynamics
  • Bandwidth
  • Control Laws
  • Robustness
  • System Integration

Video vs. Images

Need: Classroom training with realistic tools to build realistic systems

slide-6
SLIDE 6

TRIPOD: Template for Real-time Image PrOcessing Development

  • Affordable
  • Robust
  • Customizable

Input – Output Window

  • Teach Computer Vision!
  • ANSI C/C++
  • Avoids compiler specifics
  • Pointer to pixel data
  • Open-source
  • No royalties
  • TRIPOD and QCSDK are free
  • $50 USB camera
  • Logitech SDK
  • Frame rate and Video

On-line Tutorial: http://www.boondog.com/

  • Windows Compatibility
slide-7
SLIDE 7
  • 1. Open Visual C++
  • 2. Copy TRIPOD template
  • 3. ANSI C/C++ on pixel data
  • 4. Compile
  • 5. Execute and see results

Steps

Philosophy: Learn Computer Vision - Not Compiler

Pentium 3 – 500 MHz – 256 MB RAM – LEGO Camera

Hardware Specs Used:

slide-8
SLIDE 8

Project Presentation 10 Visual-Servoing Project 9 Image-based Visual-Servoing 8 Pattern recognition: SSD Tracking 7 Kernels and edge detection 6 Color recognition and tracking 5 TRIPOD exercises: threshold, binary, centroids 4 Centroids, areas, clustering 3 Memory handling: negatives, thresholds, binary, brightness 2 BMP Files, raster data, row-column vectors, pointers 1 Topic Lecture

Intro to C Programming, Linear Algebra

Classroom Implementation: Envisioned Syllabus

Computer vision fundamentals namely segmentation, pattern recognition, edge detection and tracking for robotic systems Provide students the fundamentals of computer vision. Emphasis is on understanding the physical principles governing image processing and applying them to solve engineering problems Objectives: Pre-requisites: Description: Schedule:

(3 Hrs/Week)

Text: Myler, H,. The Pocket Handbook of Image Processing Algorithms in C (ISBN: 0-13-642240-3)

slide-9
SLIDE 9

Test Case: Independent Study

Visual-Servoing Project

Image analysis, edge detection Pattern recognition, SSD tracking

http://www.pages.drexel.edu/~weg22/tutorials.html

slide-10
SLIDE 10

Test Case: Independent Study

TRIPOD: edge detection

slide-11
SLIDE 11

TRIPOD: How it Works

QCSDK (Low-level)

TRIPOD QCSDK

  • USB Interface
  • DirectX
  • Logitech API

TRIPOD (Hi-level)

  • Pointer to data
  • ANSI C/C++
  • Minimal MFC
  • Hooks to API
slide-12
SLIDE 12

Project: Wireless Imagery for C2 Augmentation C2 CENTER

Image retrieved from C2 Processed Image (edge enhancement)

PDA SCREEN VIEW

Live view

Wireless 802.11b

slide-13
SLIDE 13

Wireless Imagery for C2 Augmentation

Video Demo

slide-14
SLIDE 14

Project: Biomimic Human Head-Eye Motion for Servoing

Two Rotational DOF Camera Point-of-View

slide-15
SLIDE 15

Contributions

  • Free and open source package
  • Uses widely available and affordable cameras
  • ANSI C/C++ with minimum MFC
  • On-line tutorials and code
  • Integrate LEGO Vision Command with Mindstorms

Conclusions

  • Suitable for Independent and Honors Study
  • Can affordably and effectively implement in classroom
  • Rapidly ascend CV learning curve
  • Apply concepts to prototype robot-camera systems