Pattern recognition and crowd analysis

The present collection is the result of a review process done on the contributions received in response to an open call for a thematic special issue dealing with pattern recognition and crowd analysis. Part of the submitted contributions are significantly extended versions of papers discussed at the first edition of the Workshop on ''Pattern Recognition and Crowd Analysis'' of the, grouping contributions from the mayor experts in the field of crowd analysis and simulation. Over the last decade a growing interest in the creation of innovative computational approaches to model, simulate and analyze movements and behaviors of crowds and pedestrians involves several scientific areas and related computer-based technologies, including Computer Vision and Complex Systems Simulation. The development of novel models, solutions and technologies requires to consider the knowledge of studies and results coming from behavioral sciences, and a strict collaboration in order to create realistic and more efficient solutions to face growing requirements coming from urban planning, evacuation, surveillance, and crowd management. The investigation field is wide and requires the development of cross-disciplinary attitudes. Despite the fact that different research and application fields focus on the complex phenomena of crowd and pedestrian dynamics, many research ideas have evolved independently within vertical scientific communities. The creation of fruitful events allowing scholars and researchers coming from different scientific community to share their knowledge on this topic is a future-driven challenge to design new conceptual and technological scenarios to face needs and expectations in the understanding and management of crowds. The video analysis of people in crowd is of straightforward importance, to understand single and social behaviors, to detect anomalies and suspicious events or objects in crowds scenes, to define first-aid and crisis support in areas where big events are organized (sports, music, politics, religion, exhibitions, and so on). Experimental data are often collected by human observations and manual data collection only, to support modeling and simulation of crowds and pedestrians. These simulations are fundamental for engineering and architectural studies of new urban environments , for crisis-management avoiding panic and dangerous situation , but also for handling design-for comfort of people leaving and walking in large populated areas. The integration of these approaches will impact on the direction of this field of study and in the creation of innovative technological scenarios. Moreover, crowd analysis is becoming an emerging point of different research fields; in pattern recognition, after the stable results of the last …