Now Video Analytics has the brains to match
Bosch Camera Trainer
Built-in machine learning capabilities to train the camera
The next logical step in high-end video security, our new Camera Trainer technology harnesses “machine learning” to enable the built in Intelligent Video Analytics in Bosch cameras to function in ways that are even more accurate and application specific. A pioneering innovation that means customers have an unprecedented opportunity to turn informative data into better security, improved business efficiencies and valuable revenue screens.
Bosch Camera Trainer
Video Analytics has transformed today’s security cameras. No longer passive systems that simply observe, record and play back events, cameras can now use metadata to add sense and structure to recordings – and tell the difference between false alarms and potential threats. But while video analytics substantially improves levels of security and business intelligence, we did not believe that standard rules for detecting and reporting on security situations went far enough. It’s the reason we have spent a great deal of time and effort in developing “machine learning” technology that can be applied to our cameras. The result is Camera Trainer, a features that enables the “brains” of our security systems to be trained in ways that improve their ability to pay attention, learn, and solve specific, customized problems.
The next step in video analytics
Camera Trainer can instruct Bosch cameras to identify new objects and situations within a scene that are specifically defined by the user. Instructions can be added at any time, and are retained and used by the camera to process future scenes. These instructions can also be combined with the pre-determined alarm rules and filters that are already embedded into Bosch Intelligent Video Analytics for even greater flexibility and accuracy.
Data-driven insights for every application
But Camera Trainer doesn’t just improve movement detection, it can also recognize and detect stationery objects when they are present, or absent. And that provides far more informative data that can be used in ways that extend far beyond security.
For example, in retail environments, cameras can be taught to recognize the shape of a shopping basket or the wheel of a trolley. This enables them to detect when someone enters a fitting room or area where trolleys are not permitted, triggering an alarm to significantly reduce shop lifting.
In car parks, security cameras enabled with Camera Trainer can detect whether bays are occupied or empty, without the need for ground sensors. This saves on costs and optimizes traffic flow. What’s more, because it can identify stationary objects, Camera Trainer can also be used to detect cars that overstay their permitted parking period in time-restricted zones – a simple, yet effective way to enforce regulations and boost revenue streams.
Vehicle counting is another application that is increasingly important for smart cities. With Camera Trainer, municipalities can benefit from a highly-reliable counting system that distinguishes between in-coming and out-going traffic. So traffic signals can be adjusted to change vehicle priorities, preventing traffic jams and optimizing traffic flow in the city.
In commercial and industrial warehouses, Camera Trainer can be used for detecting loading bay status by recognizing a physical mark on the ground to detect when there is no truck in the bay. The next truck can then be sent for unloading to increase efficiency and save on costs.
Camera Trainer can also optimize warehouse logistics. By setting an occupancy alarm on conveyor belts, it can detect when there are too many boxes on a production line. This can avoid goods piling up to eliminate potential backlogs that will result in costly downtime.
In transport environments like airports, where traffic flow awareness is critical to safety and security, Camera Trainer can collect data on pre-defined situations, such as how long an airplane is parked at a terminal, or the time it takes to unload luggage or refuel. All of which can speed up turnaround time, improve efficiencies and reduce airline costs.
Camera Trainer also comes into its own for applications beyond the capabilities of standard video analytics. For example, in extreme climates, icicle formations on buildings can cause damage – and pose a threat to pedestrians. A pan-tilt-zoom (PTZ) camera can be trained to configure alarm rules to detect icicles so they can be removed in the early stages of formation, eliminating building damage and potentially saving lives.
And these are just some of the applications that can benefit from using Camera Trainer as the brains of their security and intelligence solution.
The ultimate in smart security
There’s no doubt that Video Analytics has driven innovation in security solutions. But with the introduction of Camera Trainer, it will truly prove its worth. For system integrators, it is a flexible way to tailor security to a customer’s individual needs. For end users, it’s a digital eye on their world with the brains to match. But for the world of security, it’s a whole new ball game. The peace of mind that smart data can go beyond protecting people and property to deliver business intelligence that ultimately improves efficiency and performance.
And that’s something that, until now, no security camera had the brains to match.