Вы находитесь на странице: 1из 49

Introduction to Machine Vision

Samantha Frost
Product Marketing Manager, Vision

Agenda
What is machine vision?
Why use machine vision?
Machine vision applications
Types of machine vision systems
Vision system components
Machine vision terms

Improving vision system performance


Summary

Q&A
2 | 2015 Cognex Corporation

What is machine vision?

A simple concept
Formal definition:
Machine vision is the use of devices for optical noncontact sensing to automatically receive and interpret an
image of a real scene in order to obtain information
and/or control machines or processes.
- from the Society of Manufacturing Engineers
or more simply:

Capturing an image from an industrial camera for


inspection or process control of manufactured products.

4 | 2015 Cognex Corporation

Evolution of machine vision

Proprietary Box

Embedded
Boards
PC Vision
Modular
Vision

Barcode
Readers

1982

5 | 2015 Cognex Corporation

Vision
Sensors
3D Displacement
Sensors

Today

Simple examples
Machine vision systems analyze images
Then makes decisions and/or returns numeric results
about each image it gets
Good Oil
Filter all
holes are
open

The center tab on this bracket is


37.255 mm wide

6 | 2015 Cognex Corporation

Reject oil
filter some
holes are
blocked

Image analysis
The primary purpose of machine vision is image
analysis

Decision

Image

Answer

Location

7 | 2015 Cognex Corporation

Poll question #1
What is your level of machine vision experience?
a. Im completely new to machine vision and have
never used it before

b. Im a beginner with limited machine vision


experience
c. Im an intermediate user, familiar with different
types of machine vision technology
d. Im an advanced user, but theres always more to
learn

8 | 2015 Cognex Corporation

Where is machine vision used?

Electronics/Computer
Industry

Automotive Industry
Semiconductor
Industry

Food Packaging
Industry

Consumer Products

Medical/Pharmaceutical

Shipping/Transportation

9 | 2015 Cognex Corporation

Graphic Arts/ Packaging

Why use machine vision?

A critical role in achieving strategic goals


High speed production lines
Clean room environments
Hazardous environments
Microscopic inspection
Closed-loop process control

Robot guidance
Precise non-contact measurement

11 | 2015 Cognex Corporation

Benefits of machine vision


Strategic Goal

Machine Vision Applications

Higher Quality

Inspection, measurement, gauging, and


assembly verification

Increased Productivity

Repetitive tasks formerly done manually are


now done by MVS

Production Flexibility

Measurement and gauging / Robot guidance /


Prior operation verification

Less Machine Downtime and Changeovers programmed in advance


Reduced Setup Time
More Complete Information
and Tighter Process Control

12 | 2015 Cognex Corporation

Manual tasks can now provide computer data


feedback

Benefits of machine vision


Strategic Goal

Machine Vision Applications

Lower Capital Equipment


Costs

Adding vision to a machine improves its


performance, avoids obsolescence

Lower Production Costs

One vision system vs. many people / Detection


of flaws early in the process

Scrap Rate Reduction

Inspection, measurement, and gauging

Inventory Control

OCR and Identification

Reduced Floorspace

Vision System vs. Operator

13 | 2015 Cognex Corporation

Poll question #2
What is your principal objective for attending this
webinar?
a. I am looking for a basic introduction to machine
vision
b. I want to know what kind of problems machine
vision can solve
c. I want to know the basic components of a vision
system and understand how they work together

d. I want to know which basic vision concepts I need


to master
14 | 2015 Cognex Corporation

Machine vision applications

Think: GIGI

Guidance

16 | 2015 Cognex Corporation

Inspection

Gauging

Identification

Guidance
Determines part position (x, y, and angle)
Automates handling of parts for machines:
Alignment & Placement
2D & 3D Picking
Eliminates need for fixturing & improves robot
flexibility

Vision tool alignment, fixturing


Locate at least one feature on a part for the
purpose of calculating the (x, y) position and
rotation of the part to position other vision tools
precisely

*First step in every vision application is to


find the part or features of interest
17 | 2015 Cognex Corporation

Geometric pattern matching

Trained Part

Out of focus

Confusing Background

Scale Change
Dim Lighting

180 Rotation
Reversed Polarity

18 | 2015 Cognex Corporation

Occlusion

Part location challenges


Appearance changes

19 | 2015 Cognex Corporation

Part location challenges


Presentation / Distortion effects

20 | 2015 Cognex Corporation

Inspection
Broad category of vision applications:
Correct location
Orientation
Skew

Quality
Defect Detection
Surface Inspection
Contaminants

Completeness

Fill Level
Feature Presence
Counting
Assembly Verification

21 | 2015 Cognex Corporation

Gauging
Precise dimensioning
Automated metrology and data recording

Ensure tolerances
Diameters, Gaps, Bushings, Threads, etc.

22 | 2015 Cognex Corporation

Identification
Read codes
Bar codes & 2-D Matrix
Labels & direct part mark

Read characters
OCR / OCV

Recognize objects
Based on color, shape, or size

23 | 2015 Cognex Corporation

Types of machine vision systems

1-D machine vision


100% continuous web
inspection and classification
Uses line-scan cameras
Materials inspected include:
Metals
Non-wovens
Plastics
Paper

25 | 2015 Cognex Corporation

2-D machine vision Area scan


640

480

768

1200

26 | 2014 Cognex Corporation

1024

1600

2-D machine vision Line scan


Image is built line by line
Movement is needed
Requires encoder to track movement
Short exposure times

Built Image

Line Acquired
In-Sight
5604

Line Light

Conveyor
Belt
Movement
Encoder Shaft

27 | 2014 Cognex Corporation

Four reasons to use Line scan


1. Unwrap cylindrical objects for
inspection

2. Add vision to space-constrained


environments
3. Meet high-resolution inspection
requirements
4. Inspect objects in continuous
motion
28 | 2014 Cognex Corporation

3-D machine vision

29 | 2015 Cognex Corporation

Machine vision system components

Key parts of a machine vision system


Light Source

Display

I/O
Comm.
POWER

Lens

Camera

Vision Software

A -B

SLC50/4CPU
RUN
FORC
FLT
DH+
BAT
RS232
RUN REM PROG

INPUT
0 4
1 5
2 6
3 7

INPUT
0 4
1 5
2 6
3 7

O UTPUT
0 4
1 5
2 6
3 7

ALLEN -BRADLEY

Vision System

Inputs/Outputs: switches, PLCs, robots, lights

31 | 2014 Cognex Corporation

Putting it all together

Parts
Light Source

Lens

Camera

Monitor
Input:
Serial
Parallel
ISA, PCI, VME bus
Vision System

Operator Pointing Device


(Trackball, Mouse, Touchscreen)
32 | 2015 Cognex Corporation

Output:
Serial
Parallel
ISA, PCI, VME bus

How the system works


1. Part arrives at inspection
station

2. Sensor detects part and sends a trigger to the vision system


3. Strobe is flashed to
illuminate part

FAIL!

4. Vision System acquires the image from the sensor.

5. Software algorithms running on vision system


performs image processing and/or image analysis
on acquired image

6. Vision system sends signal along a discrete output line


which activates a diverter if the part is bad
7. Operator can view rejected parts and ongoing statistics on display,
and can take system off-line if necessary

33 | 2015 Cognex Corporation

Machine vision algorithms

Transforms raw numbers into


useful higher level features
Raw numbers turned into edges, colors,
characters, and other characteristics

then makes decisions


Returns answers such as position, similarity,
distance, presence, quality

All within just tens to hundreds of milliseconds!


34 | 2014 Cognex Corporation

Poll question #3
What would you say are the principal challenges to
implementing successful machine vision on your
assembly/production line?
a. The management of my organization does not realize the
value of machine vision
b. Our line moves really fast and we have not found a vision
system that can keep up with it
c. We have variable lighting and/or environmental conditions
that make it difficult to capture a consistent image
d. The conditions of the attributes we need to analyze (e.g.,
codes, markings, etc.) are often so degraded that we
cannot read them
e. None of the above

35 | 2015 Cognex Corporation

Machine vision terms

Vision definitions
Field of View (FOV)
The part which can be
seen by the machine
vision system at one
moment. The field of view
depends from the lens of
the system and from the
working distance between
object and camera.
5 x 7

37 | 2014 Cognex Corporation

Field of view
What is my field of view?

4
8
15
38 | 2014 Cognex Corporation

Working distance and resolution


Working Distance (WD)
Distance between the FRONT of
the lens and the target

Resolution
The minimum feature size of
the object under inspection.
10

39 | 2014 Cognex Corporation

Working Distance and FOV are Interrelated

40 | 2014 Cognex Corporation

Depth of field
Depth of Field (DOF)
The distance in front of
and behind the object
which appears to be in
focus

More DOF with a small aperture


(high f/number) than with a
large aperture (low f/number)

41 | 2014 Cognex Corporation

Improving vision system performance

Different lens same image?

43 | 2014 Cognex Corporation

Lighting
Depth of field
Working distance
Line speed
Exposure time
Aperture
Budget

Lighting techniques
Light can be structured in different ways. Angle and direction of the light
determine how the mark and space is seen by the camera.

Back Lighting

Bright Field

44 | 2015 Cognex Corporation

Axial Diffuse

Diffuse Dome

Structured

Dark Field

To learn more about choosing the


right lighting go to Cognex.com >
Resources > White Papers and
Articles
4

Optics

To learn more about choosing the


right optics go to Cognex.com >
Resources > On-Demand Webinars

45 | 2015 Cognex Corporation

Filtering techniques
Control quality and quantity
of light
Block all unwanted
ambient lighting
Pass only the output of
lighting used for inspection
Increase contrast
and resolution
46 | 2015 Cognex Corporation

For more on optical filters, go to


Cognex.com > Resources > OnDemand Webinars

Image pre-processing tools

To learn more about image pre-processing, go


to Cognex.com > Resources > On-Demand
Webinars
47 | 2015 Cognex Corporation

Summary
Make products better, faster, and less expensively
Gauging, Inspection, Guidance, and Identification are
the key applications
Key components are the lens, camera, lighting, vision
software and communications
Learn more about optics, lighting, filters and image preprocessing to improve system performance
Dont go it aloneget help from experts!

48 | 2014 Cognex Corporation

Thank you for your time