Syed Babar Ali School of Sciences and Engineering (SBASSE), Department of Electrical Engineering presents
Object Based Classification Using eCognition for UAV/Satellite Data
Date: February 23-25, 2018
Time: 9:00 am to 5:00 pm
Venue: Syed Babar Ali School of Sciences and Engineering (SBASSE), LUMS
Prof. Dr. Mian Muhammad Awais
Dr. –Ing. Ahmad Kamal Nasir
The objective of this three-days workshop is to provide participants, who are planning to use Unmanned Aerial Vehicle (UAV) or satellite imagery for classification purpose, an opportunity to learn object based classification using eCognition. This workshop is useful for professional and students who are working on remote sensing projects related to agriculture, forestry, infrastructure and construction.
This workshop starts with an introduction to remote sensing with a special focus on UAV technology. The introduction sessions are followed by a UAV demo, which demonstrates the aerial remote sensing data acquisition workflow. The next session provides an introduction to classification which includes understanding of spectral signatures in a multi-spectral image for analysis purpose. Furthermore, during this session classification types and assessment techniques shall also be discussed. The classification session shall be followed by a hands-on lab session about the introduction and usage of eCognition. The next hands-on sessions demonstrate the application of advanced object based classification methods and their accuracy estimation in eCognition.
The workshop hands-on sessions includes tasks which participants shall perform under guidance. This workshop is ideal for professionals who wish to learn and implement object based classification for their projects, for teachers who wish to add UAV/Satellite data to prepare their students for remotely sensed data and for students who wish to equip themselves with working knowledge and know-how about UAV data and state of the art data classification techniques.
Course Learning Outcomes
At the end of this workshop you are expected to have the following:
• Understanding of various classification algorithms.
• Usage of eCognition for the application of object based classification on image data.
• Understanding of accuracy assessment for various classifications.
The course provides an interactive learning environment in which participants shall get opportunity for a live UAV data acquisition demo and hands-on lab sessions to perform classification on real datasets. Presentation hands-out shall also be given to the participants.
This course requires that the participants preferably have some prior knowledge about basic remote sensing and working knowledge about GIS software such as QGIS/ArcGIS or any other.
The course tuition fee is non-refundable and must be paid in advance for necessary arrangements. Because of the practical nature of the course, the course is limited to a maximum of 20 participants. The registration will be on first come first serve basis and all the interested participants have to pay the fee till Friday, February 16, 2018.
Students PKR 10,000
Professionals PKR 15,000
Note: Participants will inform through email about the accommodation requirements. Secondly, Accommodation cost is not included in registration fee.
People Interested to take this course can register either by
• Send us a bank draft in the name of “Lahore University of Management Sciences” at Electrical Engineering department, School of Science and Engineering, Opposite Sector U, DHA, Lahore Cantt, 54792. Lahore, Pakistan.
• Pay in cash to Mr. Affan Anwar, Senior Officer, Electrical Engineering department, LUMS.
Because of the practical nature of the course, the course is limited to a maximum of 20 participants
Day-Session - Topic
1-1 Introduction To Remote Sensing - Important concepts
1-2 Aerial Remote Sensing - Unmanned Aerial Vehicles
1-3 Drone (UAV) Demo
2-1 Classification, Image interpretation, EM spectrum, Spectral Signatures, Pixel based VS Object based classification & Assessment Techniques
2-2 Introduction to eCognition - Overview
2-3 Getting started with eCognition, Quick Map mode, Nearest Neighbor Classification
3-1 Advance eCogniton Usage, Rule Set Mode, Nearest neighbor classification, Classification based on indices.
3-2 Accuracy Assessment, Assessment, Export Results