Cristina Vicente Chicote

REMM-Studio: an Integrated Model-Driven Environment for Requirements Specification, Validation and Formatting.

In order to integrate requirements into the current Model-Driven Engineering (MDE) approach, the traditional document-based requirements specification process should be changed into a requirements modelling process. To achieve this we propose a requirements metamodel called REMM (Requirements Engineering MetaModel) which includes the elements that should appear in a requirements model (requirements, stakeholders, test cases, etc.) together with the relationships that may appear between them. This metamodel is the basis of the REMM-Studio environment which enables: (1) to build graphical requirements models, (2) to validate them against the metamodel and against a set of additional OCL constraints, and (3) to automatically generate a navigable Software Requirements Specification (SRS) document as the main deliverable of the Requirements Engineering process. REMM-Studio is expected to ease the integration of requirements with other development models (e.g. component models) and to facilitate the validation of the SRS thanks to its navigability.

A Model-Driven Approach for Documenting Business and Requirements Interdependencies for Architectural Decision Making

In business and system requirements analysis, the characteristics of the elements identified are perfectly documented with specific notations and in specific artifacts. However, since these notations are focused on each specific type of element, the interdependencies between elements of different kinds are usually left implicit. These relationships are particularly important during system design in order to evaluate the impact of each requirement and to select the architectural pattern that better satisfies them. To identify and make them explicit, the architect has to analyze all the artifacts generated in depth. Any misinterpretation of these relationships may lead to patterns being selected that can hinder rather than facilitate the satisfaction of the business goals and the system requirements. This paper presents a set of profiles allowing designers to explicitly model these interdependencies in BPMN 2 and UML 2 Use Case diagrams. In addition, ATL transformations are defined to automatically derive these relationships from the business specification to the requirements models, facilitating their analysis by the architect, and thus reducing the risk of misinterpretation.

Heterogeneous COTS product integration to allow the comprehensive development of image processing systems

Image processing techniques are applied in a wide range of products. Automated visual inspection of industrial products, medical imaging or biometric person authentication are only a few examples. In order to process the great amount of data contained in images highly complex and time-consuming algorithms are needed. Furthermore, many of these applications require real-time performance making specific hardware devices indispensable. Currently, there exist several Commercial Off-The-Shelf (COTS) component libraries that help to implement these hybrid software/hardware systems. In addition, some powerful tools are available that allow prototyping and simulating image processing applications prior to their implementation. However, none of these tools allows to realistically coprototype and co-simulate both software and hardware simultaneously. This work presents a new approach to the development of image processing applications that tackles the question of how to fill the gap between design and implementation. A new graphical component-based tool has been implemented that allows building image processing applications from functional and architectural prototyping stages to software/hardware co-simulation and final code generation. Building this tool has been possible thanks to the synergy that arises from the integration of several preexistent software and hardware COTS components and tools.

Face Detection on Still Images Using HIT Maps

We present a fully automatic solution to human face detection on still color images and to the closely related problems of face segmentation and location. Our method is based on the use of color and texture for searching skin-like regions in the images. This is accomplished with connected component analysis in adaptatively thresholded images. Multiple candidate regions appear, so determining whether each one corresponds or not to a face, solves the detection problem and allows a straightforward segmentation. Then, the main facial features are located using accumulative projections. We present some results on a database of typical TV and videoconference images. Finally, we extract some conclusions and advance our future work.