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

Module: Analysis of high resolution aerial data

Dates: Mon, Jun 20 - Thu, Jul 14

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:

Readings

The basic Idea of OBIA

A short introduction in Object Based Image Analysis

The basic Idea of Deep Learning

A short introduction in Deep Learning

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

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).

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.

Cloud Compare

3D point cloud and mesh processing software

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.

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.

Manual and CSF based Classification of Ground Points with CloudCompare

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

CANUPO train and classify point clouds

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

qCanupo classification

qCanupo classification tutorial

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.

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.

Assessments

Creating Orthoimages and Point Clouds

Outcome(s) assessed: