Nasim Sajadi. Outline 2 Machine Vision Machine Vision Library Methodology Taxonomy EvaluationRecommendation.

Capabilities of Machine Vision LibrariesNasim Sajadi

2

Outline

3

What is Machine Vision

Aim : Simulate human vision ability

Action: Analyse image information

Requirement: Hardware, Software, and Cameras Combination of

mathematics computer science artificial intelligence (AI) electronics

Limitations : Dependency on the image quality

4

Machine Vision vs. Computer Vision

Computer Vision

Research focus

Machine Vision

Industrial Engineering focus

5

Machine Vision in Industry

Repetitive Defect recognition

Machine Vision in Industry5

Repetitive Defect recognition

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching Measuring

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching Measuring

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching Measuring

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching Measuring

Continues Monitoring

Machine Vision in Industry5

Repetitive Defect recognition

Precise Matching Measuring

Continues Monitoring

Vision Technology Library6

HALCON7

Machine Vision

MVTec Software GmbH

Comprehensive

Operators in C++, C, C#, Visual Basic and Delphi

HALCON IDE: HDevelop and HDevEngine

OpenCV8

Open source computer vision library me

Started by Intel

C/ C++

Linux, Mac OS X and Windows ksk

Compatible with IPL & IPP

Research & Industry

Sherlock9

Machine Vision

Teledyne DALSA

Windows-based

Versions

Essential Professional

Uses MVTools library

Methodology10

Taxonomy

Extracting concepts & algorithms from documentations

Evaluation

Taxonomy >> Coverage (depth & breadth) Documentation >> strong

Good Taxonomy11

Good Taxonomy is

Comprehensive simple easy to understand and apply

Taxonomy12

TAXONOMY

Taxonomy13

Taxonomy13

Taxonomy13

Taxonomy14

Taxonomy14

Taxonomy15

Taxonomy15

Taxonomy16

Coverage of Algorithms (Low Level)

17

Edge Detectction

Image Analysis

SmoothingFiltering

Calibration

0

10

20

HALCONOpenCVSherlock

Coverage of Algorithms (Intermediate Level)

18

Segmentation

Line Extraction

3D Reconstruction

Identification

Blob Analysis

1D Measuring

Contour Processing

Morphology

0

5

10

HALCONOpenCVSherlock

Coverage of Algorithms (High Level)

19

Pattern Matching

Pattern Recognition

Motion Recognition

Face Recogniition 0

5

10

HALCONOpenCVSherlock

Documentation20

HALCON OpenCV Sherlock

Installation

Concepts & Algorithms

Access

Support Commercial Forum / Wiki Commercial

Recommendations21

HALCON OpenCV Sherlock

Vision Expertise

Programming -

Support Commercial Forum / Wiki Commercial

Task Complexity

Cost $$$ Free $$

Time

Conclusion & Future Work22

What we did

Taxonomy Evaluation

Future Work

Speed Code quality Correction

Questions??23