From Pixels to Processes

Welcome to the first ever Mid-Term Symposium of the new ISPRS Technical Commission 7, on “Thematic Processing, Modelling and Analysis of Remotely Sensed Data”. The ISPRS now has two full technical commissions dedicated to the subject of Earth Observation, our TC7 and TC8 on “Remote Sensing Applications and Policies”.

Since the formation of our new Technical Commission, the various Working Group chairs, co-chairs, scientific secretaries and regional coordinators have been busy organising various conferences and workshops around the world, and especially in preparing the Technical Programme for this Mid-Term Symposium.

At the Symposium, we have 24 Technical Sessions on the eight themes of our working groups, two plenary sessions, various special sessions, three poster sessions, opening and closing sessions as well as a series of high-level pre-symposium Workshop/Tutorials.

The symposium digital proceedings includes all the scientific/technical papers received before the publication deadline, and give an excellent overview of the status of remote sensing science and methodology. The papers on fundamental physics and modelling show that there is an increasing understanding of the basic biophysical and biochemical principles that underpin many remote sensing applications. In the various sessions on information extraction from SAR data, many novel approaches to SAR signal processing, as well as robust ways to produce interferometric SAR products have been presented. Other new trends reflected in the papers on SAR are the increasing use of polarimetric SAR information, and on high resolution SAR imagery. The three sessions on hyper spectral data contain many interesting papers on new application and processing methods being developed around the world. The two topics of the Symposium where most papers were presented are “Advanced Classification Techniques” and on the processing of multi-temporal data and change detection. With the increasing availability of extensive time serious of earth observation data, many new methods of extracting useful information are being developed.

The papers on remote sensing data fusion reflect that the trend to problem solving is in using information from multiple sources to increase reliability, robustness and confidence of the results of remote sensing data processing. The many papers on innovative problem solving methodologies for Less Developed Countries show that even with limited computer hardware/software availability, many problems can still be tackled by using earth observation data. Finally, the papers produced by the ISPRS Inter-Commission WG on “Derivation for global data, environmental change and sustainability indicators” give a good overview of the status and challenges in developing operational global and regional land observations.

On behalf of the Scientific Programme Committee, I wish you an enjoyable stay in Enschede and hope you will find the proceedings (and social events) fruitful and stimulating.

Professor John van Genderen
President, ISPRS TC7
8 May 2006

Preface to the Proceedings

From pixels to processes – the topic of this symposium stimulates us to consider the activities required to turn digital imagery into description of processes. The last Commission VII meeting in Enschede in 1986 emphasized the processing and classification of single images, understanding image properties, and linking imagery with GIS ancillary data. Remote sensing has moved on to consider processes – time series of images are being processed to provide information in a dynamic mode. Spatial scientists are being asked to deliver information relevant to current political and social phenomena, requiring new approaches to acquiring and handling significantly increased volumes of data.

In preparing these proceedings, we find there are five stages of handling, processing and understanding images:

1. obtaining and handling myriad images with increasing flexibility – in particular considering image geometry as well as spatial, radiometric, spectral and temporal resolution;
2. correcting for radiometric distortions to yield reflectance;
3. correcting for geometric distortions  in order that images may be projected together;
4. mapping cover and extracting objects at a snapshot in time;
5. understanding how features change over time – in other words understanding processes on the land’s surface, atmosphere and/or water bodies.

The papers in this volume span the full range of moving from a pixel to a process. Few cover the entire range of these steps, but a number of papers show progress is being made.

Image acquisition requires new approaches to obtaining information necessary for understanding processes. A number of promising technologies appear in these proceedings, including for example tools to handle the storage and distribution of images over the internet, flexible and low-end data acquisition solutions such as internet protocol cameras (such as web-cams), and information extraction from low-cost digital cameras.

In terms of understanding images, an exciting development is unmanned airborne vehicles. Unmanned airborne vehicles (UAVs) and low-cost satellites, both focused on providing image data at a low-cost and near real-time to users, feature in a number of papers. In 1986, virtually all work was orientated around general purpose sensor systems, such as Landsat, SPOT or NOAA AVHRR. There has been a move towards sensor systems collecting imagery and data for specific applications, such as disaster monitoring. Image sensors for process-specific tasks appear to have the benefit of higher flexibility (and sometimes lower costs) for specific applications. In other words the data needs of end-users are pushing space agency engineers to develop specific sensor systems, rather than building and operating general purpose systems such as Landsat or SPOT.

As editors, we indulge ourselves with a little crystal ball gazing of where the Commission VII conference may be focused in 20 years time. We see a multitude of cheap sensor networks located throughout the environment and linked via the internet and wireless connections.  Satellites (and UAVs) will provide imagery that give synoptic information and an ability to integrate the ground-based sensors in real-time.  With the rapid increase of data volumes, new methods for data mining and image understanding will be targeted to specific (tailor-made) digital products.

We would like to thank the local organizing and scientific committees for their input and collaboration in producing these proceedings.

8 May 2006

Scientific Program Committee

• Prof. Dr. Ir. M. Molenaar, Rector, Professor in Geoinformatics and Spatial Data Acquisition (ITC, The Netherlands)
• Prof. Dr. J.L. van Genderen, President ISPRS TC 7 (ITC, The Netherlands)
• Prof. Dr. A.K. (Andrew) Skidmore, Vice-President ISPRS TC 7 (ITC, The Netherlands)
• Prof. Dr. M. Hale, Research Coordinator (ITC, The Netherlands)
• Dr. N. Kerle, Co-Chair WG VII/7 (ITC, The Netherlands)
• Prof. A. Peled, President ISPRS TC 8, University of Haifa, Israel)
• Prof. Dr. M. Schaepman, Chair WG VII/1 (Wageningen University, The Netherlands)
• Dr. M. Inggs, Chair WG VII/2 (University of Capetown, South Africa)
• Prof. Dr. F.D. van der Meer, Chair WG VII/3 (ITC, The Netherlands)
• R. Arbiol, Chair WG VII/4 (ICC, Spain)
• Prof. Dr. Gong Jianya, Chair WG VII/5 (Wuhan University, China)
• Prof. Dr. Zhang Jixian, Chair WG VII/6 (Chinese Academy of Surveying and Mapping, China)
• Dr. O. Kufoniyi, Chair WG VII/7 (RECTAS, Nigeria)
• Prof. Dr. C. Schmullius, Chair Intercommission WG VII/IV (Jena University, Germany)

Local Organising Committee

• Prof. Dr. J.L. van Genderen (ITC, The Netherlands)
• Mr. R. Brinkman (ITC, The Netherlands)
• Mr. J. Duim (ITC, The Netherlands)
• Mr. F. Gollenbeek (ITC, The Netherlands)
• Mrs. J. Kalf (ITC, The Netherlands)
• Mrs. S. Tempelman (ITC, The Netherlands)
• Mrs. J. Terlouw (ITC, The Netherlands)
• Mr. B. Masselink (ITC, The Netherlands)
• Mr. J. Duim (jr) (ITC, The Netherlands)