548-0146/01 – Advanced methods in Remote Sensing (PMDPZ)
| Gurantor department | Department of Geoinformatics | Credits | 5 |
| Subject guarantor | prof. Ing. Jiří Horák, Dr. | Subject version guarantor | prof. Ing. Jiří Horák, Dr. |
| Study level | undergraduate or graduate | Requirement | Compulsory |
| Year | 2 | Semester | winter |
| | Study language | Czech |
| Year of introduction | 2021/2022 | Year of cancellation | |
| Intended for the faculties | HGF | Intended for study types | Follow-up Master |
Subject aims expressed by acquired skills and competences
The objective is to learn student how to utilize digital image remotely sensed data, how to pre-process data and modify image data to enhance required information, using basic as well as advanced methods of classification of digital images, including object-oriented classification and deep learning methods, processing radar and lidar data, and critically evaluate and interpret reached results.
Teaching methods
Lectures
Tutorials
Summary
The subject introduces the methods for digital processing of remotely sensed imagery. Explanation of concepts data corrections and transformations, methods of image segmentation, image filtration and using edge detectors, methods of image transformation into other coordinate systems, utlization of textural measures, pixel-based and object-based classification, soft classifiers, image spectroscopy, radar and lidar data processing.
Compulsory literature:
Recommended literature:
Way of continuous check of knowledge in the course of semester
Students are asked about knowledge from areas that they should have already known from previous lectures. Students also work on individual tasks. Tasks are frequently based on understanding of previous, simpler tasks.
E-learning
Other requirements
No additional requirements are imposed on the student.
Prerequisities
Subject has no prerequisities.
Co-requisities
Subject has no co-requisities.
Subject syllabus:
1. Digital image data from Remote sensing. Paradoxes of digital images, principles of segmentation.
2. Review of physical properties, spectral characteristics of landscape objects and phenomena and identification methods .
3. Preprocessing of digital images. Rectification methods. Radiometric and atmospherical corrections. Radiometric errors in data and its elimination. 5S model, Modtran, ATCOR 1-4, Sen2Cor, reflectivity calculation on the Earth surface, calculation of surface temperature.
4. Image enhancement methods. Thresholding, contrast modification, density slices, colour synthesis. Image filtration. Convolution. Filter separability. Low frequency filters, directional smoothing.
5. High frequency filters, edge operators, Laplacian operators, Canny detector, edge detection, pattern detection, edge delocalisation.
6. Texture, local textural measures (Haralick function), image segmentation methods (based on thresholding, edge detection, region based, hybrid methods), detection of geometric features, Hough transformation.
7. Object-based image analysis (OBIA). Utilization of segmentation, methods of delimitation of image objects, algorithms Baatz-Shäpe. Fourier transformation.
8. Spectral indeces. Variants of vegetation indeces, indeces for mineral detection, humidity, snow and others. Data fusion for images with different spatial resolution
9. Pixel-based clasification. Supervised spectral clasification of multispectral images. A training phase, correction of training sites. Parametric and non-parametric classifiers.
10. Pixel-based unsupervised classification methods. K-means, ISODATA, ISOCLUSTER. Hybrid clasification. Neural networks and advanced techniques of classification (deep learning, convolutional neural network).
11. Soft classifiers, utlization of Bayes theorem, Dempster-Shafer theory, classification and uncertainty. Assessment of classification results. Post-classification modification.
12. Image spectroscopy, hyperspectral and ultraspectral data. A review of sensors. Preprocessing and atmospherical corrections, flat field conversion, empirical line, models, PCA, MNF. Pixel purity index. End members and their discovering. Classification of hyperspectral data. Applications.
13. Processing data from radar systemms. Principles, biasesFactors influencing resulting signal. Coregistration of a pair of SAR products, interferogram creating and estimation of coherence, removal of demarcation from interferogram, image filtering. DInSAR method. Radar polarimetry. LIDAR and UAV. Processing of lidar data. Application fields.
Conditions for subject completion
Occurrence in study plans
Occurrence in special blocks
Assessment of instruction