Machine Vision Algorithms 

LEVEL : ADVANCED      HRDF : CLAIMABLE

Trainer: Dr Eric Ho Tatt Wei

 
WHEN

5 - 7 JULY 2021
 
WHERE

MS TEAMS (ONLINE)

 
TIME
9.00AM - 5.00PM

COURSE SUMMARY

INTRODUCTION

The course teaches the key concepts and working principles of state-of-art machine vision algorithms. The course material covers all major classes of algorithms that are required in a practical and industrial machine vision workflow to perform complex automated inspection tasks. The course focuses on the modern probabilistic algorithms approach which has been demonstrated to give robust and repeatable performance in industrial and practical settings. 

Autonomous systems are the key agents that will implement Industry 4.0. More complex automated inspection tasks which are higher in the manufacturing value chain almost always require machine vision algorithms to direct operations of robotic and manipulator systems. This course addresses the critical need in Malaysian industries to upgrade their machine vision implementations towards modern probabilistic approaches which has consistently demonstrated greater robustness, reliability and higher performance on complex automated inspection tasks. The course aims to help engineering employees in Malaysian companies bridge the technology gap by imparting a deep and intuitive understanding of the key concepts and working principles underpinning modern probabilistic machine vision algorithms. The goal is to develop sufficient technological sophistication and versatility in modern algorithms to enable Malaysian companies to transition their machine vision algorithms and products towards more complex inspection tasks and move up the automated inspection value-chain. 

COURSE CONTENT
Fundamentals of digital images 
  • Sources of information content. 
  • Sources of variation and noise. 
  • 3D to 2D projection, loss of information, depth of field, perspective and orthonormal views. 
  • Pixel level features and descriptors: local features with scale and affine invariance.
  • Pixel level probabilistic features: case study of binary classification of background subtraction.
  • Machine vision as a classification problem.

Image Preprocessing and denoising

  • Whitening/Wiener filtering, Histogram equalization.
  • Bilateral filter, LARK filters, BM3D, iterative filtering.   

Realistic local-pixel models 

  • Multivariate probabilistic models & hidden/latent variables.
  • Mixture of Gaussian, Factor Analysis, t-distribution models.
  • Expectation Maximization Algorithm.  
  • Case study for segmentation. 

Algorithms for identification 

  • Probabilistic subspace models – Gaussian mixtures & Probabilistic Linear Discriminant Analysis.
  • Case study of object identification from known sets & for clustering.

Algorithms for shape recognition 

  • Shape templates.
  • Statistical shape models. 
  • Probabilistic PCA subspace shape model.

Algorithms for temporal object tracking 

  • Kalman filter.
  • Particle filter.   

Higher order vision models for complex object recognition 

  • Graphical vision models : markov random fields & belief networks.
  • Case study of image segmentation with Markov Random Fields & Gaussian Mixture Models. 


WHO SHOULD JOIN US TO :

Target Position 

  • Junior and Senior R&D Engineers, Product Development Engineers

Target Sector/Industry 

  • Automated Test Equipment Manufacturers, Semiconductor and Electronic Board and component manufacturing and test companies, automotive inspection.  

Target Location 

  • Penang, Perak, Kuala Lumpur, Melaka, Johor, Sarawak. 


OBJECTIVES

Upon completion of this course, participants will be able to:

  • To strategize, select and justify state-of-art machine vision algorithms for task-specific automated inspection based on fundamental principles. 
  • To adapt and integrate several state-of-art machine vision algorithms to perform a series of complex tasks in automated inspection systems. 
  • To estimate and assess the performance of state-of-art machine vision algorithms for task-specific automated inspection systems.





OUR TRAINERS



1. Dr Eric Ho Tatt Wei

Dr Eric Ho Tatt Wei received his MS and PhD degrees in Electrical Engineering from Stanford University in Silicon Valley, USA specializing in computer hardware and VLSI systems, As part of his PhD research, he developed real-time systems for fruit flies for biological research to conduct automated inspection and guide robotic manipulation. He is currently pursuing applications of deep neural network technology to network analysis on MRI brain images.


COUNTDOWN

DaysHoursMinutesSeconds

REGISTRATION FEES

      ProfessionalS       

myr2,250

*fee quoted does not include SST, GST/VAT or withholding tax (if applicable)

Early bird/ group/ student

myr2,025

*fee quoted does not include SST, GST/VAT or withholding tax (if applicable)

OUR LOCATION

Centre for Advanced & Professional Education (CAPE)

 Level 16, Menara 2, Menara Kembar Bank Rakyat, 50470, Jalan Travers, Kuala Lumpur.

CALL US

+605 - 368 7558

DROP US AN EMAIL

cape@utp.edu.my