Software training

Introduction to typical machine learning tasks and algorithms

• Overview of problems that can be solved with the help of machine learning methods
• Overview of various classification algorithms:
  • Support vector machines (SVM)
  • Multilayer perceptron (MLP)
  • Gaussian mixture models (GMM)
  • k-nearest neighbors (k-NN)

Solving image processing tasks with machine learning methods

• Data acquisition and data preparation
• Feature extraction and selection
• Training and inference flows in HALCON
• Solving the following tasks:
  • Segmentation
  • Anomaly / defect recognition
  • Image classification
  • OCR

Segmentation

Segmentation

Classification

Classification

Defect detection

Defect detection

Advanced topics

• Regularization
• Automatic feature selection
• Hyperparameter optimization
• Interpretation of various performance metrics

Training objectives

• Evaluation of image processing tasks with regard to their resolution via machine learning methods.
• Selection of the most suitable machine learning method for your image processing task
• Data acquisition and data preparation for machine learning training procedures.
• Autonomous solving of classification / segmentation and defect detection tasks with the help of machine learning methods in HALCON

Training topics are accompanied by practical exercises. For the exercises, please bring your own laptop to the training session.

Course fees (per person)

The courses include drinks, lunch and participation certificate. For first-time HALCON customers, the course fee is €450.00 + VAT per person; for existing HALCON customers, the fee is €350.00 + VAT per person. Upon the purchase of a HALCON SDK (up to six months after training), first-time HALCON customers will receive a one-time €100 discount on the SDK purchase price.

Deep Learning in HALCON

• Overview of deep learning
• Deep Learning vs. traditional machine learning methods
• Background knowledge about deep learning in industrial machine vision applications
• Data structures for deep learning in HALCON

Classification with Deep Learning

• Acquiring and labeling of data for classification
• Programming of one's own deep learning image classification application
• Analysis of a deep learning image classification

Object Detection with Deep Learning

• Introduction to the MVTec Deep Learning Tool
• Introduction to object detection with deep learning
• General workflow for object detection

Semantic Segmentation with Deep Learning

• Introduction to labeling data for semantic segmentation
• General workflow for semantic segmentation

Anomaly Detection with Deep Learning

• Introduction to anomaly detection in HALCON
• General workflow
• Specific parameters for anomaly detection

Additional topics

• Integration of pre-trained networks in ONNX format
• Visualization of intermediate layers of networks

Object detection

Segmentation

Anomaly detection

Classification

Course fees (per person)

The courses include drinks, lunch and participation certificate. For first-time HALCON customers, the course fee is €450.00 + VAT per person; for existing HALCON customers, the fee is €350.00 + VAT per person. Upon the purchase of a HALCON SDK (up to six months after training), first-time HALCON customers will receive a one-time €100 discount on the SDK purchase price.

Appointment

March 24, 2022, from 9 am - 4 pm CET, online, in English.
Please register by March 10, 2022. If there are fewer than five participants in total, the event will not take place.

Introduction to the MERLIC user interface (MERLIC backend)

• Inserting and connecting tools
• Working with EasyTouch
• Working with triggers
• Parallel processing

MERLIC designer

• Overview of the widget library
• Design of several frontends as tabs or dialogs
• Setting up the user administration

Development of applications in the MERLIC backend

Creating a graphical user interface with MERLIC Designer

Defect detection

Work with 3D height images

MERLIC frontend

• Starting up the frontend
• Administration of different user groups
• Multiple frontends
• remote frontends

Standard applications and workflows

• Get to know the basic MERLIC tools
• Example applications such as counting, measuring and defect detection
• Work with 3D height images

MERLIC process integration

• Setting up process integration
• Creation of recipes
• Defining of interface parameters

Course fees (per person)

The courses include drinks, lunch and participation certificate. For first-time MERLIC customers, the course fee is €450.00 + VAT per person; for existing MERLIC customers, the fee is €350.00 + VAT per person. First-time MERLIC customers will receive a one-time €100 discount (up to six months after training) on the purchase price.