This course aims to teach students how to integrate close-range 3D computer vision with synthetic aperture radar satellite remote sensing. The course covers topics such as image measurements, semantic analysis, point cloud data reduction, segmentation, and synthetic aperture radar remote sensing. Students will learn skills such as camera orientation, feature extraction, noise influence analysis, and performance comparison of segmentation algorithms. The teaching method includes lectures, demonstrations, and practical projects. This course is intended for individuals interested in computer vision, remote sensing, image processing, and research in related fields.
Overview
Syllabus
Integrating Close-Range 3D Computer Vision and Synthetic Aperture Radar Satellite Remote Sensing
TUB Computer Vision & Remote Sensing Research Group
Medical Image Processing Projects
3D-Reconstruction Tool Chain
Discovering How Images Cover Object Graph of image triplets • Descriptor variance similarity to identify loop
OpenOF • Framework for sparse non-inear least squares optimization on a GPU
Image Measurements and knowledge: Camera Orientation Relative to Rectangle
Semantic Analysis • Various disciplines require semantic analysis of point clouds
Simplification - MDL Implies Trade-Off and fitting the measurements? . MDL principle: Transmitter wants to encode the information
Point Cloud Data Reduction Using MDL Principle
Segmentation into Low-Curvature and High-Curvature Segments
Voronoi-Based Extraction of a Feature Skeleton 11
Noise Influence on Point Cloud Segmentation
Performance Comparison of Point Cloud Segmentation Algorithms
Synthetic Aperture Radar (SAR) Remote Sensing
SAR Raw Data Processing Matched Filtering
Processing of SAR Data
Differential Interferometric SAR Phase Difference = f(Movement)
Compressive Sampling! Sparse Signal Representation Signal f can be transformed to a domain where it is sparse using an orthonormal basis eg a wavelet transform
Sparsity: Images in Wavelet Domain
Compressive Sampling II
Multibaseline SAR Acquisition
Proposal for Future Research: Use of Ubiquitous 3D Reconstructions
Taught by
UCF CRCV
