Wednesday, November 25, 2009

New background image for the header

This weekend I spared some time to finally play with the background image of the header.
The algorithm was this: I put random rectangles on the terrain (increasing the height of its points by 1) until somewhere the height reached 255. (The result was more satisfying than I first thought.)
After this I tried to create the lines.
The first try was a simple modulo, if the height at a given point is the multiple for instance 25, there will be a gray point on the map. This is the result:

As you can see, it's like a galaxy.. not a map. The lines are not really connected. The shape can be seen, but still...
So, the next approach was: check the neighbours, and if there is a difference in height by modulo 25, we'll draw a gray point.
Result:

Much better! But the lines are too thick.. an edge detection will solve the problem: you'll only check a certain point, and only one of its neighbours: the uppermost.
Result:
And the final version, which we are using now as a background:
It makes the small text quite unreadable, but at least we have something like a map. I'll play some more with this, maybe I'll use only an image, putting on it red or blue text, so it will be more readable.

Tuesday, November 24, 2009

GeoServer, GeoTools and GeoBatch: supporting operational Meteorology and Oceanography

Found this one today, I don't have the time now, so I cannot tell whether it's worth watching or not:
http://blip.tv/file/2799458

Seattle Conference on Scalability: MapReduce Used on Large Data Sets

Google Tech Talks June 23, 2007 ABSTRACT 2007 Google Seattle Conference on Scalability:

"Using MapReduce on Large Geographic Datasets Speaker: Barry Brumitt, Google, Inc. MapReduce is a programming model and library designed to simplify distributed processing of huge datasets on large clusters of computers. This is achieved by providing a general mechanism which largely relieves the programmer from having to handle challenging distributed computing problems such as data distribution, process coordination, fault tolerance, and scaling. While working on Google maps, I've used MapReduce extensively to process and transform datasets which describe the earth's geography. In this talk, I'll introduce MapReduce, demonstrating its broad applicability through example problems ranging from basic data transformation to complex graph processing, all the in the context of geographic data."

Video here:

Sunday, November 22, 2009

Full-waveform ALS workshop - Exercise C - Classification and filtering of full-waveform ALS data

Third exercise. Last step of this workflow: Classification and filtering. Here we learned some interesting things about OPALS and SCOP++ software's. First time, when I used SCOP++, but I think, it is an useful software to create DTM's....

Classification and filtering of full-waveform ALS data from Gottfried Mandlburger.

"Program:

Perform the following tasks with one of the prepared datasets:

1) Digital Surface Model (DSM) using OPALS
a) Import first echoes of FWF dataset (project_first_echo.xyz ) into the OPALS Data Manager (opalsImport)

b) Calculate a Digital Surface Model (opalsGrid) and quality models (sigma, excentricity)

c) Visualize the resulting DSM as hill shading (opalsShade) and color coded raster map (opalsZColor)

d) Visualize the sigma and excentricity model (opalsZColor) and compare and interpret the results visually with respect to the hill shading

2) Analysis of Full Waveform (FWF) Attributes:
a) Import last echoes of FWF dataset (project_last_echo.xyz ) into the OPALS Data Manager (opalsImport)

b) Perform raster analysis of FWF attributes (opalsCell):
• Amplitude
• Echo width

c) Derive color coded visualizations of the attribute/echo width raster (opalsZColor)

d) Compare and interpret the results visually with respect to the hill shading

3) Perform a standard DTM filtering / classification of the point cloud
a) Derive a Digital Terrain Model based on the last echoes (project_last_echo.xyz) using SCOP++ (Robust Interpolation, strategy :Lidar DTM Default)

b) Visualize the DTM (hill shading, Z-Coding, Isolines …)

4) Filtering of the point cloud with pre-classification on echo width basis
a) Repeat the steps 3a and 3b for the last echoes with small echo widths (ew<1.9ns (project_last_echo_small_ew.xyz)

b) Compare and interpret the results visually with respect to the results of Step 3 by means of the hill shading

5) Extra task: Difference model
a) Derive a normalized Surface Model: nDSM = DSM-DTM (opalsDiff)

b) Visualize the nDSM (opalsZColor)"

Full-waveform ALS workshop - Exercise B - Quality control and georeferencing

Quality control and georeferencing are the next steps in the workflow.

Second exercise - Quality control and georeferencing from Camillo Ressl

"Geometric accuracy – motivation

• Accurate georeferencing isr equired for properly using the ALS data:
- 2 components: absolute georeferencing & relative georeferencing

Absolute georeferencing of the data (checking it always requires some sort of external ground control data)

Relative georeferencing: describes the orientation of the ALS strips with respect to each other. How well do the strips fit together? Check of consistency and accuracy, compute difference between neighbouring ALS-strips."

We learned here, how can we check the georeferencing..... I write later more about this theme!


Full-waveform ALS workshop - Exercise A - Waveform Analysis

The exercises were the best part of the workshop. I think, the work in small groups (eight people) was very beneficial.




First Exercise: Waveform Analysis from Thomas Melzer and Andreas Roncat

"In this exercise, full-waveform analysis is demonstrated with a sample data
set of the 2005 ight mission over the Schonbrunn area, Vienna. In the rst
section of the exercise, single waveforms will be analysed, the second part focuses
on the overall processing of this sample data set."



Second part. Work with OPALS.

Exercise A-2: Fullwave Analysis Using Opals from Thomas Melzer

"This exercise illustrates how to use OPALS to
- perform fullwave analysis on the output of a RIEGL LMSQ-560 scanner (SDF format)
- transform the resulting point cloud from the scanner system into project coordinates, e.g. UTM ("weak" georeferencing)."



I wolud like to give some information about OPALS:

"OPALS stands for Orientation and Processing of Airborne Laser Scanning data. It is a modular program system consisting of small, well defined components referred to as ’modules’. Each module is accessible in three different ways:

• as commandline executable from within a DOS/unix shell
• as Python module from within a Python shell
• as C++ class via DLL linkage

Modern ALS sensors provide a huge amount of data (several Mio. points per km²). Thus, efficient management of point cloud data is a precondition for sucessful processing of ALS projects. Within OPALS, the administration of point and line related geometries is based on the OPALS Data Manager (ODM). The ODM acts as a spatial cache and features:

• high performance spatial queries (nearest neighbour and range queries) using two different spatial indices (K-d-tree for points and R*-tree for all other objects)
• an extensible adminstration scheme for storing arbitrary attributes (additional infos like (eg. echo width, echo number, amplitude, classification, surface normal vector, etc.) on a per point basis." (source: OPALS_gettingStarted.pdf)

more here: http://www.ipf.tuwien.ac.at/opals/opals_docu/index.php

I think, in the future we will hear more about this excellent software research....

Friday, November 20, 2009

Michael Jackson's 3D Body Scans on eBay for $1.5 Million

That was a great idea in 1996. However, I am astonished by it since 3d scanners were not too wide spread in the 1990's.

Link: http://gizmodo.com/5407507/michael-jacksons-3d-body-scans-on-ebay-for-15-million



Wishful thinking; Imagine there is a company to scan recent singers and other stars for movies and production in the future.

Wednesday, November 18, 2009

Full-waveform ALS workshop - Experience with operational FWF ALS - Projects in Archaeology

Ninth Presentation

Michael Doneus: Experience with operational FWF ALS - Projects in Archaeology

CONTENTS:

- Archaeological background –Aerial Archaeology
- ALS andVegetation
- Technical Issues
o Types of Scanning Systems
o Filtering
o Georeferencing
o Acquisition -Time Frame
- Archaeological Issues
o Interpretation
o ComparisonwithTerrestrial Survey
o ALS andArchaeological Prospection
- ArchaeologicalApplications

CONCLUSION:

Only if we understand the technology, issues and limitations coming with data collection, filtering and interpretation we will be able to successfully apply ALS within various disciplines.

Full-waveform ALS workshop - Breaklines and DTM filtering with FWF ALS

Eighth Presentation

Christian Briese: Breaklines and DTM filtering with FWF ALS


Overview:

• Motivation
• Structure line extraction
- Automated structure line modelling
- Automatic start segments
• DTM generation
- Standard methods
- Improvement by full-waveform (FWF) ALS
• FWF attributes
• FWF data management
• Improved DTM generation based on an echo width threshold
• Extended robust interpolation by individual a priori weights determined from FWF
attributes





Summary:
Improved DTM determination with FWF ALS

• Additional Information available by FWF ALS data
Per echo: distance, amplitude, echo width, cross-section
• Interesting results in order to detect last echoes reflected by low vegetation
• Further studies are still necessary:
- Analysis of the influence of the footprint size, the incidence angle, …
- Analysis of the accuracy and reliability of the FWF attributes (distance, echo
width, amplitude, …)
- Comparison of different sensors
• Extension of the Filtering and Classification Methods
- additionally to the typically purely geometric criteria – additional FWF echo
attributes are available and should be integrated into the classification process
- however, a large area based detailed analysis of the advantages of the FWF
attributes is necessary in the future

Full-waveform ALS workshop - Signature analysis of full-waveform ALS data

Seventh Presentation

Werner Mücke, Wolfgang Wagner: Signature analysis of full-waveform ALS data



Overview:

- Brief revision
- Additional full-waveform (FWF) parameters (directly observed or derived)
- Interaction of the laserbeam with the surface
- Effect of different surface types on the FWF observables
- Scale / typical values of FWF observables
- How can FWF observables be used operationally?
- Examples
- Usage of FWF observables (e.g. point cloud classification, vegetation
detection, tree-species identification, structural information, ...)




Summary:

- Introduction to FWF ALS data
- Further understanding of interaction of the laser beam with different
surface types
- How can FWF ALS measures be used operationally?
- Examples for
Tree species classification
Retrieval of structural information
Point cloud classification

Full-waveform ALS workshop - Georeferencing of ALS Data

Sixth Presentation

Norbert Pfeifer: Georeferencing of ALS Data


Overview:

• Multi Sensor System
• Errors
- Data driven point of view
- Components of the system
• Mathematical models for correction
- Model driven approaches
- Approaches not considering the measurement process
• Observation of discrepancies
• Examples
• Summary / Points of further discussion




Summary:

• Elevation- and planimetric-coordinates of the “original” points on the ground and upon objects are
affected by systematic errors caused by the GPS/INS/Laser scanner system. 􀃆 This leads to
erroneous georeferencing. (direct georeferencing)
• Errors may be static or dynamic and global or varying from strip to strip.
• First quality check: Computation of differences in the strip overlap
• Elevation in height alone is not sufficient
• 3D correction of exterior and interior orientation (calibration) is necessary
• Solution: ALS strip adjustment with rigorous mapping function (polar measurement and trajectory from
GPS/INS with tie patch parameters)
- Minimization of errors at autoamtically determined homologous objects (planes)
- Quality parameters (improvement of std.dev. After adjustment)
• For high precision demands (relative and absolute accuracy) ground control is required.
• Flight planning must consider cross strips and control point areas.
• Quality control is a must for each block! Graphic display of the discrepancies is a thorought check for
detecting systematic errors (excluding the datum).
• Future aims: Standardization in ALS
- GPS/IMU preprocessing
- Fligh parameters: strip overlap, flight patterns (cross strips)
- Quality of ALS data should be estimatable a priori (as for image-based AT)

Full-waveform ALS workshop - Radiometric Calibration of Full-Waveform Small-Footprint Airborne Laser Scanner Data

Fifth Presentation

Christian Briese, Hubert Lehner: Radiometric Calibration of Full-Waveform Small-Footprint Airborne Laser Scanner Data






Conclusions:

• Radiometric calibration essential for adequate use of full-waveform ALS
sensors
• Practical methods have become available
• Further R&D
- Radiometric accuracy (in %)
- Internal calibration (monitoring of sensor functions)
- Deviations from radar equation
- Object signatures at 1 μm, 1.5 μm and other ALS frequencies

Full-waveform ALS workshop - Waveform Analysis

Fourth Presentation

Andreas Roncat, Thomas Melzer: Waveform Analysis

Scattering Cross-Section:

• A single scatterer is characterized by its differential scattering crosssection,
σ(t) which depends on its
- size
- reflectivity
- roughness
- directionality and solid angle of reflectance (BRDF)
• The measured echo (return signal) due to a scatterer depends also on the
- incidence angle between emitter/scanner and surface
- original strength of the emitted pulse
- attenuation of the pulse due to distance (R-square-law) and occlusion


Full-waveform ALS workshop - OPALS

Third Presentation

Gottfried Mandlburger, Johannes Otepka, Wilfried Karel: OPALS Software – Orientation and Processing of Airborne Laser Scanning data

new software concept, very great research!!!


Software Concept:

• Modular structure composed of small, well defined units (modules)
• Availability of modules as:
- Command line programs
- Python modules
- C++ classes via DLL linkage
• Individual process control via scripts
- Shell scripts (Unix/Linux), Batch (MS Win2000/XP/Vista)
- Python
• Management of point cloud data based on the OPALS Data Manager (ODM)
• Interfaces for efficient data interchange with DTM-, CAD-,GIS- and
Visualisation/Modelling-Software
- SCOP++, …
- AutoCAD, MicroStation, …
- ArcGIS, Grass-GIS, Quantum-GIS, …
- 3D-Studio, GeoMagic, Deep Exploration, SMS (Surface Modelling System)
• Use of standard and open-source libraries (boost, GDAL/OGR, CGAL, ...)
• Abdication of interactivity

Full-waveform ALS workshop - Principle of Full- and Online Waveform Analysis in Airborne, Mobile and Terrestrial Laser Scanning

Second presentation was from Riegl Company. I already wrote about Riegl VZ-400 scanner, here we heard some Online Waveform Analysis applications....

Dr. Andreas Ullrich, DI Peter Rieger: Principle of Full- and Online Waveform Analysis in Airborne, Mobile and Terrestrial Laser Scanning

Overview:

- Introduction
- Laser Scanner RIEGL LMS-Q680 for Full Waveform Analysis
- Full Waveform Analysis and Online Waveform Processing
- Full Waveform Sample Data
- Multiple Time Around Technique
- RIEGL V-Line Laser Scanners with Online Wavefrom Processing
- Calibrated Reflectance
- Multi Target Capability
- Online Waveform Sample Data


Making 3D maps on the move

http://www.technologyreview.com/computing/23967/?a=f

Full-waveform ALS workshop - Measurement Principle and Physical Fundamentals

First presentation

Wolfgang Wagner: Measurement Principle and Physical Fundamentals

Overview:

• How does a full-waveform laser scanner work?
• Important properties of laser light
• Range determination and range resolution
• Beam pattern and spatial resolution
• Radar equation
• Cross section of different targets
• Waveform generation
• Waveform analysis → Presentation of Andreas Roncat
• Radiometric calibration → Presentation of Christian Briese



Conclusions:

• Full-waveform laser scanners depict the measurement process in its
entire complexity
- Full-waveform system are compatible with ranging systems, but not vice-versa
• Advantages
- Algorithms can be adjusted to tasks
- More echoes as in first/last pulse systems
- Intermediate results are respected
- Neighbourhood relations can be taken into account
• Calibration of the data, i.e. conversion to cross section, is essential for
physical modelling efforts
• The additional data, i.e. amplitude, width, cross section, is valuable for
segmentation and classification purposes
- Classifying terrain and non-terrain points for DTM filtering

Full-waveform ALS workshop successfully completed

Hallo everybody,

Thank you, workshop was great, good organized. DTM generation is a big topic!!! I learned some new information about full-waveform....we see here a picture from Vienna...


Saturday, November 7, 2009

14. International Congress "Cultural Heritage and New Technologies" - Vienna

TerraFormatics will be there! You will find some fresh news on our blog!




Workshop 14 will be held
from November 16th to November 18th 2009 in Vienna, CITY HALL (Rathaus der Stadt Wien)

MAIN TOPIC 2009:
ARCHIVING - or building an information system

Archiving is today central to nearly all aspects of Cultural Heritage Management.
Archives (Archaeological excavations, libraries, documents, data collections,...) are important data repositories.
The data contained in correctly treated and accessible archives makes wide and varied information available.
How can archiving in all its aspects best promote knowledge about and support the protection and conservation of cultural heritage?

Link:

http://www.stadtarchaeologie.at/

Digital Terrain Models from Full Waveform Laser Scanning Workshop

TerraFormatics will post some information about this workshop!


organizers:

Univ.Prof. Wolfgang Wagner
Univ.Prof. Norbert Pfeifer

Institute of Photogrammetry and Remote Sensing
Christian Doppler Laboratory for Spatial Data from Laser Scanning and Remote Sensing
Vienna University of Technology
___________________________________________________________________________
Date: November 12 - 13, 2009
Venue: Vienna University of Technology-Gusshausstrasse 27-29-1040 Wien-Austria
___________________________________________________________________________



Motivation
Airborne laser scanning is a technology which has developed rapidly in the last few years. It has set new standards, especially for the generation of digital terrain models (DTM). A great advantage of laser scanning is the ability to “see” through gaps in the vegetation foliage. The DTM quality strongly depends on the correct classification of the 3D point cloud into terrain and off-terrain echoes. Especially dense and low vegetation is considered as problematic. With full waveform laser scanners, commercially available since 2004, a considerable amount of these problems can be overcome. Additionally, new quality indicators can be derived.

Goals
This workshop will teach the basics of full waveform laser scanning in lectures, exercises, and discussions. It will reach from the sensor specifications to terrain modeling. The following list of topics will be treated.

* Measurement principle and physical foundations
* Sensor properties and measurement process
* Waveform analysis and calibration
* Geo-referencing
* Signature analysis of full waveform parameters
* Classification and segmentation
* Filtering (classification) based on full-waveform information
* DTM derivation
* Quality control

Three practical exercises will be carried out, next to a demonstration of a full waveform sensor. Each participant will follow each exercise in groups – with a maximum of ten persons per group.

Wednesday, November 4, 2009

digital 210 king

A very good blog about Building Information Modeling (BIM) with Autodesk applications.


http://www.digital210king.org/blog.php?p=5&elq=nl

Shazam

Something similar is going on with feature extraction of scanned data. This is just audio:
(In laser scanning this technique is used to map point clouds - taken from different angle - together.)

DuckDuckGo

http://duckduckgo.com/?q=laser+scanning&v=d
A new search engine: with 'zero click', you'll get results about the searched topic, so you can avoid clicking the results until you find some useful information.
Pretty impressive.