INNOVATION > SITUATIONAL AWARENESSPresenting Different Approaches to Scene Change DetectionEmbedding greater intelligence into surveillance systems remains a priority for security as well as safety and operational monitoring objectives. However, with technologies such as video analytics not always living up to expectations when deployed into the real world, understanding how to adapt and apply the right techniques in a given situation is critical. Digital Barriers is a world leader in video analytics and surveillance intelligence, creating solutions that can be rapidly deployed as part of enterprise surveillance systems, as well as developing and configuring solutions to meet more complex and specific requirements. Some video analytics approaches rely on simple pixel based approaches to detect changes in scenes. In contrast, we apply a number of techniques, including zone and rule based detection based on a number of detection parameters, as well as background learning based approaches to detection. This provides more options for real world situations.
What you can see below are applications of our scene change analysis technologies to detect left objects. These reflect our ability to adopt a range of approaches, including both readily deployable commercial solutions and more sophisticated bespoke solutions. SCENARIO 1: Rule Based Approach to Scene Change Detection
The below example demonstrates scene change detection using zone and rule based models with configurable parameters. These include the 3D detection zone - including the tracking of individuals into and out of the zone, minimum and maximum object dimensions, sensitivity to the visibilty of the object in the scene, detection schedules and alarm delays.
SCENARIO 2: Background Learning Approach to Scene Change Detection
The below example illustrates scene change detection on a thermal image source. It uses background learning to detect a significant and persistent change in the scene. Key to the technique is the rapid updating of the background model to account for natural scene changes, whilst still being able to specifically detect a left object and ignore other subjects.
FURTHER INFORMATIONFor further information on our approach to video analysis and scene change detection, or for a demonstration of our technologies in action, please contact us at:
Digital Barriers Solution Engineering |
Situational SEE OVERVIEW >> |
