As more and more Earth Observation (EO) data becomes available, the need to automate at least some aspects of data processing is apparent. The SURF project was funded by the European Space Agency (ESA) to provide a survey of image-processing methods for EO and an in-depth analysis and prototyping of some of the most promising methods. The survey has included (1) a list of application areas within EO; (2) the development of criteria for the evaluation of methods; (3) a classification of image processing tasks within EO, independent of the applications; (4) single-page descriptions of a wide range of methods. Based on this background work, a dozen methods were selected for further analysis and considered for prototyping.
The next stage of the project consists in prototyping four of the methods subjected to in-depth analysis. This paper presents the results of the survey and a brief review of the methods selected for prototyping.
In addition to the substantial amounts of available Earth Observation (EO) data, there is currently an increasing trend towards the acquisition of larger and larger EO data and image quantities from single satellites or missions, with multiple, higher resolution sensors and with more frequent revisiting. More sophisticated algorithms and techniques than those largely in use today are required to exploit this rapidly growing wealth of data and images to a fuller extent. The project “Survey and Assessment of Advanced Feature Extraction Techniques and Tools for EO Applications” (SURF) funded by the European Space Agency (ESA) will address these issues. The objective of SURF is to provide an overview of the current state-of-the-art Methods within feature extraction and manipulation for EO applications and to identify scenarios and related architectures for exploitation of the most promising EO feature extraction Methods. The task is to identify the most promising Methods to extract pertinent information from EO data on environment, natural resources and security issues. SURF aims at listing existing Methods with the final goal of identifying the three most promising Methods to be implemented in prototype solutions. The work includes the development of the concept for the evaluation and rating of Methods relative to the users needs for information, the maturity and novelty of the Methods, the potential for fusing data and the operational feasibility. Special emphasis will be made regarding the exploitation of state-of-the art image processing, pattern recognition and classification techniques.