PureTech Systems has developed and patented a suite of geo-spatial video content analysis solutions called PureActiv. PureActiv integrates with ONSSI’s Ocularis video surveillance platform, providing industry leading video detection, behavior recognition and object classification. PureActiv utilizes video streams from an IP camera or encoder, analyzes the images in real time using geospatial video analytic algorithms and sends metadata alarms directly to Ocularis for display to the operator. The integration also includes tampering alarms and is deployable on all fixed and PTZ cameras with the system.
PureActiv video analytics processes video streams coming from selected cameras. Each camera can maintain its own unique detection capabilities. Detected events like intrusion, people counting or traffic incidents are sent to the Ocularis video surveillance platform using a software development kit (SDK) XML connection. Configurable alarms descriptions, meta-data and bookmarks are attached to every event for further processing. In addition to “typical” alarm meta-data, PureActiv also supplies more advanced data, such as latitude, longitude, elevation, object classification and object speed.
PureTech Systems’ PureActiv differentiators at a glance include –
- Patented Geospatial Video Detection Algorithms.
- High-performance detection with proven low false alarms rates.
- Industry leading technical support to integrators.
- Generation of robust real time metadata.
- Robust detection capability including object classification (animal, person, vehicle), real size, real speed and exact location (latitude / longitude / elevation).
- Includes camera tampering alarms.
- Applicable to both fixed and PTZ cameras.
- A proven solution for critical infrastructure, transportation, airports and utility markets.
- Low nuisance alarm rate – Multi-modal “learning” background feature.
- High detection rates in difficult, long range outdoor environments.
- Works with visible, IR illuminated and thermal cameras.
- Deployable as a server or edge solution.
PureTech Systems’ PureActiv video analytics are integrated with OnSSI’s Ocularis video surveillance platform. The integration is achieved by using the Ocularis SDK and PureActiv’s robust API. IP video is received by PureActiv over an IP connection directly from a camera or video encoder. The video stream is them analyzed for intrusion event or specific behaviors as defined in the PureActiv setup tool. When detection has occurred, alarm information is collected, packaged into standard metadata format and reported to the Ocularis server via an IP connection. The PureActiv setup tool is where regions of interest, rules, object classification and behaviors can be defined per camera. Operators can also manage several PureActiv specific operations like defining schedules, geo-referencing the scene and defining specific object classes and size. The alarms raised by PureActiv are then stored in Ocularis video surveillance platform and can be reviewed at any time.
Benefits of PureActiv include highly accurate, long range detection capability for outdoor scenes. Longer detection range and higher detection rates equate to improved security, cost savings and commercial gains. PureActiv’s metadata allows for increased efficiency in defining and reacting to alarms. The ability to use a wide range of camera models and camera types, as well as, the flexibility to install as a server or edge device solution gives the installer more options and provides further cost saving for the end user.
The ability to detect objects and intrusions using various types of cameras and illuminants while ignoring normal environmental conditions
The ability to determine "real" object size. This is in contrast to just determining pixel size. Although an object may encompass a very large number of pixels, this algorithm enables the software to compute the object's size in the real world regardless of how large it may appear in the camera view.
The ability to count vehicles with directional accuracy, while ignoring pedestrians and environmental conditions.
Detection that a vehicle, or other object of interest has stopped within a defined region.
Background modeling is crucial to every form of analytic behavior intended to be used in outdoor applications. It affords the capability to recognize traits of the background and remove those attributes from the scene. This then allows for analysis of the remaining objects.
Detection within a Region
The use of video analytics becomes more powerful when allowing the ability to add additional requirements to the behavior. For example, the entrance or exit from a defined area, or a specific object type of action occurring in that area.
The identification of objects by type: human, car, truck, boat, unknown. The subsequent classification can then be used for alarming and display.
Camera movement is commonplace, annoying and places havoc on most motion detection products. The Stabilization algorithm removes this video shifting providing a stable video for 3rd party systems and operator viewing.
Wrong Direction / Speed Threshold
This algorithm can detect the real speed of the vehicle or object, regardless of its orientation to the camera (approaching, departing, moving across scene), as well as, the direction of movement with respect to an area of interest. Thresholds can then be set to trigger appropriate alarms.
The algorithm looks for objects that have been left in a location, or have been thrown into the view of the camera.
Loitering / Crowd Detection
Often times it is important to not only detect the object, but understand how many objects are present and whether they are remaining in a defined area for a suspicious period of time. This behavior is covered by Loitering and Crowd Detection.
Video analytics can also be used to detect possible tampering with the camera. The Camera Tampering function looks for the loss of the video image (Due to coverage or disconnection) or excessive movement of the camera (Due to removal or redirection).
Video Analytic Capabilities
|Critical Infrastructure||Utilities||Mass Transit||Airports||Railroads||Military|
|Object moving into/out of area||√||√||√||√||√||√|
|People and vehicle tracking||√||√||√||√||√||√|
|People / Vehicle counting||√||√||√||√||√||√|
|Object Classification/Real Size||√||√||√||√||√||√|
|PTZ or Fixed||√||√||√||√||√||√|
Additional Information / Links
|People and Vehicle Tracking||Video|
|Object Classification/Real Size||Video|