LV-UAV-Guide | Analysis of high resolution aerial data

Outcomes

The module is intended to provide an initial insight into the automated and evaluation of high-resolution orthoimages and the point clouds. Since this topic easily fills several courses, the possibilities of more complex evaluations with QGIS and comparable software will be presented. The module offers:

A lightning introduction to and classification of Object Based Image Analysis (OBIA), Machine Learning (ML and Deep Learning (DL) in relation to the task of the course.
a template example of an OBIA classification run with QGIS and Orfeo Toolbox (OTB)
the optional implementation of DL and ML classifications with R
A first classification and analysis of point clouds with the goal to develop at least in principle in this area

Readings

The basic Idea of OBIA

A short introduction in Object Based Image Analysis

basic mandatory YouTube 5min/27min

The basic Idea of Deep Learning

A short introduction in Deep Learning

basic YouTube 1min/4min optional

A review of Object-based image analysis (OBIA) in (UAV) remote sensing

An overview of OBIA especially for land-cover mapping purposes using remote-sensing imagery

advanced mandatory

Deep learning versus Object-based Image Analysis (OBIA) in weed mapping of UAV imagery

Object-based image analysis (OBIA) and deep learning approaches were applied to the weed mapping task of the UAV imagery. The article provides comprehensive quantitative measure of the the performance of both approaches (copy in Ilias).

advanced recommended optional

An applied comparison of Machine Learning (ML) versus Deep Learning (DL) approaches using UAV Imagery

The article provides a overview of the practical impacts and pay offs applying advanced technics lie DL and ML to an applied research task.

advanced optional

Cloud Compare

3D point cloud and mesh processing software

advanced optional

Experiences

OBIA Workflow for QGIS in a nutshell

The basic Object-based image analysis (OBIA) workflow with QGIS and the OTB processing plugin follows a straightforward approach. This Video shows a very common way.

basic mandatory YouTube 18min

OBIA Workflow for QGIS step by step

The basic Object-based image analysis (OBIA) workflow with QGIS and the OTB processing plugin follows a straightforward approach. This Tutorial shows the most common way.

basic mandatory

Manual and CSF based Classification of Ground Points with CloudCompare

Manual/Automatic classification and automatic segmentation for small photogrammetric datasets using CloudCompare.

basic optional YouTube 6min

CANUPO train and classify point clouds

The CANUPO suite is a simple yet efficient way to automatically classify a point cloud.

advanced optional

qCanupo classification

qCanupo classification tutorial

advanced optional YouTube 19min

DL & ML Workflow with R

There is a complete advanced course dealing with deep learning an AI in R. Below you will find the necessary parts for deep learning only.

advanced supplemental optional

Deriving ecological indices from LiDAR data

After the simple classification of ground points, a typical processing of the data with LiDAR tools can be largely performed.

advanced supplemental optional

Module: Analysis of high resolution aerial data

Outcomes

Readings

The basic Idea of OBIA

The basic Idea of Deep Learning

A review of Object-based image analysis (OBIA) in (UAV) remote sensing

Deep learning versus Object-based Image Analysis (OBIA) in weed mapping of UAV imagery

An applied comparison of Machine Learning (ML) versus Deep Learning (DL) approaches using UAV Imagery

Cloud Compare

Experiences

OBIA Workflow for QGIS in a nutshell

OBIA Workflow for QGIS step by step

Manual and CSF based Classification of Ground Points with CloudCompare

CANUPO train and classify point clouds

qCanupo classification

DL & ML Workflow with R

Deriving ecological indices from LiDAR data

Assessments

Creating Orthoimages and Point Clouds