OVERVIEW
The participant is guided through hands on sessions alternated with exercises to recap the lessons learned. Goal: after this training the trainee has the capability to set up a basic Image analysis workflow on its own, understanding the different steps to come to the desired result on your own data and topic.

---

CONTENT
Module 1 explains basic functionalities from loading data to exporting the results. Different sensor
types are used for the hands-on exercises.
• Load and view data, Introduction to Processes
• Segmentation: How to create Image Objects; Image Objects - the information carriers
• Basic Classification; Classify using context information: Feature ‘Relative border to class’
• Merge Objects; Export results
• Sample based Classification with Nearest Neighbor classifier
• Batch-processing with eCognition Server

Module 2 teaches advanced classification methods and refinement strategies. A land cover
classification example is used analyzing a Quickbird subset.
• Using Customized Features for classification; Classification using fuzzy Membership Functions – enhance transferability
• Semantic relationship and inheritance
• Refinement of Classification using algorithms “find enclosed by” and “ remove objects”
• How to reuse a Rule Set

Module 3 explains the usage of multiple Object Levels, as well as using thematic GIS information.
Advanced Classification strategies about how to use logical operators and variables are introduced
using an example of change detection.
• Introduction to change detection using thematic layer and current classification
• Method to classify changes: Use information about sSb Objects and thematic layer
• Classify the changes using GIS information; Alternative: Translate values in “Object Variables”

---

TIME
2 - 4 March 2010
All days 9.00 am - 5.00 pm

PRICE
1.600,- Euro per seat for 3 days standard training with Definiens instructor

VENUE
Brock & Michelsen A/S
Blokken 76
DK - 3460 Birkerød