Learn about how to use the "Unsupervised Classification" algorithm (former "cluster analysis" algorithm) within the Trimble eCognition software with ease. This is based on the ISODATA (Iterative Self-Organizing Data Analysis Techniques) approach and creates a single raster output with pixel values corresponding to the cluster ID. Matthias will show you which parameters are crucial and how to implement this in your rule set. Enjoy!
Please find more information, the data & rule set for this video and discussions in our eCognition Knowledge Base: https://support.ecognition.com/hc/en-us/sections/360004561580-eCognition-Deconstructed
______________Video Content_________________
00:00 - Introduction & Theory
04:37 - Example
05:01 - Unsupervised Classification (alg.)
08:35 - Multi-threshold Segmentation (alg.)
09:25 - Create/Update Class (alg.)
12:04 - Assign Class (alg.)
13:42 - Merge Region (alg.)
15:30 - Unsupervised Classification on Image Object Level (alg.)
17:15 - Multi-threshold Segmentation on Image Object Level (alg.)
