The TOSCLA project, funded by the Community of Madrid (CCG10-UC3M/DPI-5535), is developing an advanced computer vision system specialized in detecting moving objects within complex and dynamic environments. Its primary focus is on overcoming critical challenges such as sudden changes in lighting, moving backgrounds, and occlusions, thereby ensuring accurate and reliable real-time identification. To achieve this, the system implements a methodology based on data fusion and the use of a set of classifiers (ensemble technique).

By combining information from multiple sources and processing it using various algorithms in parallel, the system unifies individual criteria to reach a robust final decision, dramatically increasing accuracy in applications such as autonomous driving, robotics, and intelligent video surveillance.