We embrace the latest open-source AI/ML science developed by Google, Facebook and Microsoft. Our solution is a unique blend of in-house technology and flexible system architecture, that is easy to maintain and engineered to expand and solve new problems. This allows us to integrate state-of-the-art research for the best result in the shortest possible time.
iSentry is designed to analyse CCTV cameras in real-time and give that information to controllers immediately in the form of video alerts that are enriched through multiple data points. From there iSentry is able to add instant visual insight to the huge amounts of footage generated by a large-scale CCTV deployment.
Any unusual behaviour of persons or objects is detected. This in turn enables the identification of behaviour such as fights, violence and aggression, hijacking, loitering, unauthorised access, trip and fall, operational irregularity, smoke and fire, traffic accidents and many more.
iSentry detects and classifies a wide variety of objects, ranging from persons, helmets, backpacks, mobile phones to birds. It produces precise data for the operator or the rules engine to be processed further.
This is possible through the video tripwire and threat ranking and extraction (TREX) modules. Even small and fast-moving objects can be detected and contextualised. This is ideal for border control, intrusion detection and public space security.
These allow the detection of specific objects such as face masks being worn correctly or not, distances between people/objects, counting of people/objects (allowing for density control), helmets or cash within banks, among others.
iSentry redefines the real-time video analytics market, providing customers with a quantum leap in reaction time, efficiency, safety, operational effectiveness, and true situational awareness.
No single technology route can ever achieve this. We have developed a globally unique product set to process, contextualise and visualise real-time video data.
The combined solution is effective because a complex but logical process sequentially analyses video. It is extremely quick, using the least amount of hardware overhead resulting in a low Total Cost of Ownership. It is available across multiple hardware/software platforms and architectures