English
REMOTE SENSING AND DIGITAL IMAGE PROCESSING
Presential
The course is structured in 4 sequential modules of 3 weeks each. The first two modules are concentrated on Remote sensing topics and the last two on photogrammetric topics. Learning outcomes are defined per module and evaluated progressively at the end of each one.
Module 1: Digital Image Processing
In this module, you will understand and apply the basic radiometric preprocessing like atmospheric calibration, spatial and temporal filtering and contrast enhancement operations, which is essential in a geospatial problem-solving process. Besides, you will explore the integration of spectral bands in indices and ratios to provide sufficient insight into the information contents of the multi and hyperspectral data sets.
Module 2: Advanced Image Classification
In this module, Random Forests (RF) classifier will be taught and used to classify both single-date and multi-temporal satellite images. Various strategies for generating samples required to train supervised machine learning classifiers and assess their classification results will be explained in detail.
Module 3: Earth Observation Sensors for Mapping Applications
In this module, you will have a comprehensive overview of airborne Earth observation sensors. The course will also treat in more focus the new platforms and sensors relevant to large-scale mapping applications, including Unmanned Aerial Vehicles (UAVs), laser scanners and mobile mapping systems. Besides, you will apply Global Navigation Satellite Systems (GNSS) receivers to measure the point coordinates and assess the quality of the results.
Finally, the Image orientation process, as one of the primary tasks of any photogrammetric procedure, will be taught. This process will result in a solution that relates the image space with the object space. Thus, the absolute or relative 3D position of an object, visible in the stereo pair, can be extracted.
Module 4: 3D Data Acquisition from Aerial Imagery
In this module, you will learn the geospatial data processing techniques to derive 3D and 2D geoinformation from a sequence of overlapping drone images. Besides, you will process drone multispectral images that can be used for agricultural applications. Furthermore, the module presents oblique aerial photogrammetry, where subjects such as the Nadir (vertical) versus Oblique, accuracy overview, image processing, and applications will be explained.
During the module, hands-on experience will be gained using the appropriate software packages to process different data sets and assess the outputs.
Upon completion of the RS modules, you will be able to:
Select appropriate sensors and image data for geospatial problem solving
Apply relevant contrast enhancement for visual and digital image analysis
Apply spatial and temporal filters to improve image data for visual and digital image analysis
Calculate indices and ratios for digital image analysis
Apply different strategies for generating training and validation samples for supervised machine learning classifiers
Apply various feature selection methods for data dimensionality reduction purposes
Summarize the main multi-temporal image analysis steps
Apply Random Forest classifier to classify both single-date and multi-temporal images
Critically interpret the classification results obtained by applying supervised machine learning classifiers
Upon completion of the photogrammetry modules, you will be able to:
Describe the UAV properties and classifications and distinguish the two main mapping applications
Describe the sensor system properties, output data quality and applications for EO sensors with a focus on laser scanner and mobile mapping systems
Explain 3D point cloud properties and data quality generated by a laser scanner, and apply basic processing methods on the 3D point clouds dataset for mapping applications
Differentiate the quality of the positional control and define appropriate required positional accuracies for various applications
Design flight planning for a specific application
Understand image orientation procedures (direct and indirect) with a focus on digital aerial images.
Apply image orientation procedures, point cloud and orthophoto generation procedures, and feature extraction procedures on drone images using the designated software.
Process aerial oblique images (image orientation, point cloud, orthophoto generation) using the designated software
Process the multispectral drone images using the designated software.
Assess the quality of all the procedures mentioned above and the produced data.
3
REMOTE SENSING AND DIGITAL IMAGE PROCESSING
NaN
Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or HaDEA. Neither the European Union nor the granting authority can be held responsible for them. The statements made herein do not necessarily have the consent or agreement of the ASTRAIOS Consortium. These represent the opinion and findings of the author(s).
