Passive people or vehicles recognition and tracking for video surveillance systems
SentiVeillance is designed for real-time biometric face identification, people and vehicles tracking, and automatic license plate recognition. Based on proprietary award-winning algorithms, the technology ensures security and enables faster decision-making in various scenarios from law enforcement, city traffic monitoring, security, and other applications.
SentiVeillance technology features:
- Multiple modalities – biometric facial recognition, people/vehicle tracking and ALPR are supported. These modalities that can be used either on their own or in combination, subject to the surveillance system design.
- Automatic operation – SentiVeillance-based system can be configured to automatically report events, such as watch-list matches, or perform automatic enrollment from video streams.
- Large surveillance system support – the technology can analyze video streams from multiple cameras in parallel on a single server or a cluster of machines.
- Lighting conditions tolerance/adaptability – optical, near-infrared and thermal imaging cameras are supported by SentiVeillance.
- Real-time performance – SentiVeillance-based system performs face recognition, pedestrian or vehicle classification, and tracking in real time, both on live streams or video files.
Two products are available:
- SentiVeillance Cluster – a ready-to-use scalable solution, designed for integration with video management systems (VMS) to enable automatic, real-time monitoring and analysis using data from multiple video streams observing vehicles and pedestrians in real-time. The solution allows to connect several machines into a cluster to process larger number of video streams. Besides real-time video analysis, the solution supports searching in the events logs via the VMS user interface using a photo or a still-frame.
- SentiVeillance SDK – a software development kit, designed for software developers who are creating surveillance systems based on SentiVeillance technology.