Real-time Complexity Analysis of Live Images

OntoView™ is a real-time monitoring system which processes live video images. The program analyzes the complexity of the captured live image content and alerts of traffic problems in real time. The revolutionary aspect of OntoView™ is that it doesn’t analyze images based on conventional image-processing techniques. Instead, it emulates the way the brain perceives images, by measuring image complexity and deducing from it the intensity of the phenomenon under observation.

Ways to Use:

A natural application of OntoView™ is city or motorway traffic, whereby the instantaneous traffic intensity and “stability” are deduced from complexity and plotted in real-time. Viewers can use OntoView to detect traffic complexity fluctuation in the monitored area, take snapshots and record the maximal or minimal complexity in history.

OntoView™ offers a new way to monitor crowds. The understanding of crowd behavior and dynamics is safety-critical in many circumstances, such as sports events or manifestations. Even though crowds are made up of single individuals, the behavior of a crowd is not simply the sum of the individuals. Collective behavior in a crowd may in fact emerge suddenly and without any apparent early-warning. However, since crowd complexity may be linked to the collective “level of anger” or “energy” in the crowd, monitoring of crowd complexity with OntoView can provide critical pre-alarms.

OntoView™ can process infrared images signaling, for example, the presence of intruders or objects which enter a given perimeter. The entrance of an intruder into a scene disrupts its complexity.

OntoView™ is easy to install and operate. It can handle web-cams or IP cameras and allows users to define and choose from multiple sources. OntoView™ captures images up to 30 frames per second from USB cameras. And the following releases will support any other video device including Analog cameras connected to capture card, TV-boards, camcorders with FireWire (IEEE 1394) interface and from Network cameras.

The OntoView™ GUI is illustrated below

Features

Supporting Multiple Cameras and Groups

multiple USB Cameras and groups
description, video quality, and video resolution
USB Cameras

Video Viewing

Supports 1 camera viewing
Viewing captured images in realtime
Shrinking or growing captured image for desired image size in per-cents of original image.
Drawing meshes on captured image in user specified color, dashstyles.

Complexity Monitoring

Viewing complexity history of the monitored area in real time
Historic playbacks of complexity by time or events
Adjustable time range frame rate over the LAN/Internet to match the real frame transmit speed, and to keep the video synchronized with the actual time

Complexity Alarm

Complexity fluctuation detection function for each camera
Event triggering such as alerts or alarms when user-specified thresholds of complexity are crossed
Snapshots taken in playback mode to get BMP images for printing or saving as evidence
Adjustable alert and alarm thresholds

Example of infrared live images analyzed by OntoView™.