Theory and Algorithms

Terms of Reference

• Development of algorithms for the analysis of remotely sensed data
• Development of automated feature and attribute extraction techniques and methodologies from multi-sensor, multi-resolution, multi-spectral, and multi-temporal imagery
• Fundamental research into image understanding and object recognition
• Integrated sensor orientation
• Image sequence analysis
• Development of algorithms for including features in the orientation processes
• Research on the spatial, spectral and temporal properties of natural and human-formed objects and their image signatures
• Analysis of characteristics of hyperspectral and multi-sensor data
• Study of the relationship between object models in object recognition and GIS

WORKING GROUPS of ISPRS COMMISSION III

• WG III/1 - Sensor Pose Estimation
• WG III/2 - Surface Reconstruction from Images as Information Source
• WG III/3 - 3D Reconstruction from Airborne Laser Scanner and InSAR Data
• WG III/4 - Automated Object Extraction
• WG III/5 - Algorithms for Industrial Vision
• WG III/6 - Multi-Source Vision
• WG III/7 - Modeling Large Scale Urban Environments
• WG III/8 - Reliability and Performance of Algorithms

WG III/1 - Sensor Pose Estimation

WG III/1  Terms of Reference

• Block adjustment: projective vs. perspective transformation
• Registration algorithms
• Orientation procedures for 3D scene reconstruction
• Block triangulation for airborne digital sensors and cameras
• Use of features as entities in image orientation processes

WG III/2 - Surface Reconstruction from Images as Information Source

WG III/2  Terms of Reference

• Stereo matching
• Shape from X
• 3D versus 2.5D (in collaboration with WG III/3)
• Improvement provided by automated object identification and by image/scene understanding (coordination with WG III/4)

WG III/3 - 3D Reconstruction from Airborne Laser Scanner and InSAR Data

WG III/3  Terms of Reference

• Algorithms for point cloud processing (in collaboration with WG III/2 on surface reconstruction)
• Data fusion (in collaboration with WG III/6)
• Products: Digital surface models, digital elevation models, 3-D city and landscape models
• Applications: (in collaboration with WGs VII/3, VII/4, VII/5 on coastal mapping, flood prediction, urban planning, telecommunications planning, monitoring of power lines, noise and gas propagation, tax verification, real estate sales etc.)

WG III/4 - Automated Object Extraction

WG III/4  Terms of Reference

• Segmentation and aggregation/grouping of image features based on classification and computer vision approaches
• 3-D object extraction and image/scene understanding (in collaboration with WG III/2)
• Knowledge representation and manipulation, control structures, management of uncertainty, and learning, i.e., automatic model generation
• Geometric, semantic and temporal modeling of man-made and natural objects like buildings, roads, and vegetation including their relations in satellite, aerial and close-range imagery
• Utilization of prior knowledge, especially in the form of CAD models, GIS, or results from digital surface model analysis
• Performance evaluation (quality control, test procedures) jointly with WG III/7 and III/8
• Liaise with the Computer Vision community

WG III/5 - Algorithms for Industrial Vision

WG III/5  Terms of Reference

• Calibration: Off-line versus on-line, geometric versus radiometric calibration; active versus passive systems, non-stereo such as depth-from-focus and other shape-from-X techniques
• Reconstruction: real-time versus non-real time, 2D versus 2.5D versus 3D, multiple view geometry versus other shape-from-X algorithms, calibrated versus uncalibrated methods
• Recognition: object-centered versus viewer-centered, quantitative versus qualitative
• Model- and appearance based inspection
• Micro-surfaces: sensor models, active image acquisition, microscopic shape-from-X
• Performance and Reliability: assessment of the investigated industrial vision algorithms, in collaboration with WG III/8 and WG V/1

WG III/6 - Multi-Source Vision

WG III/6  Terms of Reference

• Information fusion from multi-resolution multi-source data such as SAR, multi-spectral, hyper-spectral, panchromatic and laser scanner data
• Modeling of uncertainty in multi-source computer vision
• Using additional knowledge sources, such as GIS, to support object extraction
• Evaluation of the effectiveness and efficiency of multi-source information fusion

WG III/7 - Modeling Large Scale Urban Environments

WG III/7  Terms of Reference

• Integrated/simultaneous analysis of terrestrial and aerial imagery for urban model and texture extraction
• Automated generation of urban models with level-of-detail and inferred attribution
• Merging of information from remotely sensed imagery, traditional cartographic products, CAD models, and urban GIS
• Techniques for integration of GPS, automated image matching, and interactive construction of virtual environments
• Photo-realistic rendering for a moving user
• Optical tracking and navigation for augmented reality

WG III/8 - Reliability and Performance of Algorithms

WG III/8  Terms of Reference

• Data set: setting up a large source data set for algorithm evaluation with B&W and color mages, various stereo overlaps, airborne laser and InSAR reference data
• Digital airborne data source: test data sets
• Reference terrain models and digital map data
• Simulation: data sets for optical, SAR and laser sources for parameter tuning of algorithms
• Specifications: rules for the validation of algorithms and the tuning of algorithm parameters
• New quality criteria: surface shape rendering with discontinuities, slope breaks, surface roughness, quality versus compression rates of DSM triangulation techniques (in collaboration with WGs III/2 and III/7)
• Accuracy versus robustness: definition, evaluation and tradeoffs
• International algorithm comparison: establishment of evaluation protocols and organization of international algorithm comparisons