Machine Vision Introduction

jabgoldfishAI and Robotics

Oct 19, 2013 (3 years and 9 months ago)

86 views

Machine Vision Introduction

ECE5320 Mechatronics

Assignment#01: Literature Survey on Sensors and Actuators

Prepared by: William Bourgeous


Dept. of Electrical and Computer Engineering

Utah State University

E
-
mail: williamk@cc.usu.edu

Phone: (
435) 750
-
0147

Machine Vision Introduction

Machine Vision Introduction

Slide
-
2

Outline


Basic working principle


Benefits


Major applications


Basic system types


Capabilities


Limitations


Selected Examples


Selection and Purchasing


Reference list


To probe further

Machine Vision Introduction

Slide
-
3

Machine Vision


In the past, Human eyes did what no machines
could do:


locating and positioning work, tracking the flow of parts, and
inspecting output for quality and consistency.



“Today, the requirements of many manufacturing
processes have surpassed the limits of human
eyesight.”



“Machine Vision provides manufacturing
equipment with the gift of sight.”


Courtesy of:
Cognex

Machine Vision Introduction

Slide
-
4

Machine Vision Working Principle


In a typical machine vision application:



A video camera positioned on the production line
captures an image of the part to be inspected and
sends it to the machine vision computer.



The computer then uses sophisticated image
analysis software to extract information from
images and generate decisions about those
images, such as: (next)

Courtesy of:
Cognex

Machine Vision Introduction

Slide
-
5


Where is it?



Locate objects accurately even within complex or confusing
scenes.



How good is it?



Inspect objects to ensure quality and consistency.



What is it?



Identify objects by analyzing their shapes or by reading serial
numbers on their surfaces.



What size is it?



Make measurements.


Courtesy of:
Cognex

Machine Vision Working Principle (cont)

Machine Vision Introduction

Slide
-
6

Courtesy of:
dvtsensors

Machine Vision Working Principle (cont)

Machine Vision Introduction

Slide
-
7

Machine Vision System Components:

The key system components are:

1.
Lighting

2.
Camera

3.
Part Sensor

4.
Imaging

5.
Processing

6.
Inspection Software

7.
Communications/Networking


Courtesy of:
datx

Machine Vision Introduction

Slide
-
8

Machine Vision Benefits


No “fixturing” of parts required


Completely Visual, No touching or bumping


Streamline operations


Can be integrated with overall system


“Eyes get tired. People make mistakes.

I have a high degree of confidence with

machine vision systems that are

properly set up.”


-

Vincent Conforti, Donnelly Electronics

Courtesy of:
Data Translation

Machine Vision Introduction

Slide
-
9

Capabilities


Machine vision is used in various industrial and
medical applications. Examples include:


Electronic component analysis


Signature identification


Optical character recognition


Handwriting recognition


Object recognition


Pattern recognition


Materials inspection


Currency inspection


Medical image analysis


Courtesy of:
Whatis.com

Machine Vision Introduction

Slide
-
10

Limitations


Machine vision is currently limited by a combination
of the following three items:


Processing Power


Communication Lines


Camera Resolution


Integration Cost and Time

Courtesy of:
Whatis.com

Machine Vision Introduction

Slide
-
11

System Types


Triangulation


Laser Radar


Sonar


Stereo Vision (2D)


(2 and ½ D)


Hybrid Systems


3D Vision


3D Vision is the most
advanced and
applicable technology
at this time.



Laser Radar

Triangulation

Machine Vision Introduction

Slide
-
12

Object Recognition, Measurement, Condition

-
System can locate objects
based on their edge
characteristics


-
Advantages:


Orientation independent


Scale independent


Touching or overlapping
parts


Lighting variation


Repeatability


Courtesy of:
adept

Machine Vision Introduction

Slide
-
13

One Camera 3D System

1.
Algorithms use a single still image from a compact
CCD video camera mounted on the robot end
-
effector to calculate the full 3D location of the part
(i.e., x, y, z position and roll, pitch and yaw angles).


2.

This information is transmitted to the robot
controller over a high
-
speed communication line.
The robot controller uses this to guide


3.
The robot's hand and intercept each part correctly
for grasping or performing other robotic processes.



Courtesy of:
Braintech


Machine Vision Introduction

Slide
-
14

One Camera 3D System (Cont)

Courtesy of:
Braintech


Machine Vision Introduction

Slide
-
15

Typical (3D) Application

Courtesy of:
Braintech


Machine Vision Introduction

Slide
-
16

Ford Example

System Objectives:


Identify randomly placed engine block


Identify orientation with respect to robot end
effector.


Accurately position robot to “grab” part


Correctly maneuver and deposit part at
destination, possibly in motion.


Complete tasks with a Single Camera.

Machine Vision Introduction

Slide
-
17

Ford Example (cont)

Courtesy of:
Braintech


Machine Vision Introduction

Slide
-
18

PCB Connector Examination Example


Challenge:


An electronics manufacturer needed an accurate, cost
-
effective
method to inspect the orientation of connectors on a printed circuit
board (PCB).


Machine Vision Solution:


A good image showing the correct orientation of the connectors on the
PCB was stored. A picture of the production PCB is obtained via frame
grabber and compared to the original. The image passed or failed
based on the match against the good image.


Machine Vision Introduction

Slide
-
19

Getting Started

1.
Learn about the Technology


2.
Identify Application


3.
Match the Technology to the Application


4.
Identify Vendors Who Can Provide Solutions


5.
Execute a Feasibility Demonstration


6.
Write the Specification


7.
Request Quotations


8.
Evaluate Responses


9.
Purchase

10.
Integrate


11.
Maintenance

Machine Vision Introduction

Slide
-
20

Reputable Suppliers


Braintech


Cognex


Adept


Data Translation


National Instruments


Applied Machine Vision (AMV)


Machine Vision Introduction

Slide
-
21

To explore further


Automated Imaging Association AIA
http://www.machinevisiononline.org


Braintech


http://www.braintech.com


Society of Mechanical Engineers SME
http://www.sme.org/cgi
-
bin/communities.pl?/
communities/mva/industrylinks.htm&&&SME&


Intel Open Source Computer Vision Library
http://www.intel.com/research/mrl/research/opencv/




Machine Vision Introduction

Slide
-
22

Reference list


Machine Vision Online


3D Machine Vision as a Shop Floor Metrology Tool

http://www.machinevisiononline.org/public/articles/
General_Electric.pdf


Single Camera 3DTM (SC3DTM)
http://www.machinevisiononline.org/public/articles/Bab
ak_Habibi_July03.pdf


Geometric Object Recognition in Robotics
http://www.sme.org/downloads/mva/Pelton.pdf


Machine Vision Today
http://www.sme.org/downloads/mva/Mancini.pdf


Machine Vision Introduction

Slide
-
23

The End


Questions?


Prepared by: William Bourgeous


Dept. of Electrical and Computer Engineering

Utah State University

E
-
mail: williamk@cc.usu.edu

Phone: (
435) 750
-
0147