ISR

Chapter deals with MUSIS (Multinational Space-based Imagery System) European cooperation, a conventional approach to developing and acquiring high-quality space systems based on collective needs, pooling of resources, and cost-sharing. This applies equally to the defense and security domain, where comprehensive earth observation competences are operationally required. In this context, the cooperation under MUSIS will be detailed as a characteristic paradigm. MUSIS acronym stands for “Multinational Space-based Imagery System” and represents the initiative of 7 + 1 defense ministers of European Union members (Belgium, France, Germany, Greece, Italy, Poland, Spain, and Sweden in the joining) to produce surveillance, reconnaissance, and observation capabilities, as required in the time horizon of 2015– 2030. MUSIS started out ambitiously as an idea for a European Earth Observation System for defense and security, with multinational produced space components and a commonly defined and developed user ground segment. However, from conception to deployment, MUSIS has been proven a complicated endeavor and time-consuming process so far. MUSIS has received skepticism for not being able to evolve into a tangible space program yet, with finalized and validated framework for its operational deliverables, participation requirements, and resource allocation. The reasons are nested in the requirements for multinational coordination among its participants, the formalities of national and collective decision making, and the individual political, operational, and industrial imperatives at national level. MUSIS remains an open intergovernmental project with a certain political guidance, but the lack of legally binding programmatic provisions so far plus the ongoing financial crisis leaves room for ambiguity in participants’ intentions and further commitments regarding MUSIS. This chapter addresses the background upon which MUSIS is based, the satellite program description and its development, the space components architecture, and the MUSIS foreseen operation and reaches some conclusions regarding the program and its future prospects.

In integrated segmentation and recognition (ISR) of handwritten character strings, the underlying classifier is desired to be accurate in character classification and resistant to non-character patterns (also called garbage or outliers). This paper compares the performance of a number of statistical and neural classifiers in ISR. Each classifier has some variations depending on learning method: maximum likelihood estimation (MLE), discriminative learning (DL) under the minimum square error (MSE) or minimum classification error (MCE) criterion, or enhanced DL (EDL) with outlier samples. A heuristic pre-segmentation method is proposed to generate candidate cuts and character patterns. The methods were tested on the 5-digit Zip code images in CEDAR CDROM-1. The results show that training with outliers is crucial for neural classifiers in ISR. The best result was given by the learning quadratic discriminant function (LQDF) classifier.