Moving Human Detection and Counting
TOSALL focused on building systems for observing humans and understanding their appearance, movements, and activities, providing advanced interfaces for interacting with humans, and creating realistic models of humans for various purposes.
Understanding human activity has applications in various fields, the most important of which is surveillance. Other applications include character animation for games and movies, avatars for teleconferencing, advanced intelligent user interfaces, biomechanical analysis of actions for sports and medicine, etc.
First we use automatic methods for finding humans in an image or a video. Once the human is detected, depending on the application, the system can do further processing to go into the details of understanding the human activity.
We have a human detection IP that is generic and easily portable on any platform including but not limited to GPU based architectures. The software can be easily ported in CUDA or openCL based on the end platform.