The practicality of advanced security video analytics software, and why you should care.
Since the late ’90s, video analytics security systems have received a range of reviews from “This is the greatest advance in security technology in 50 years” to “It doesn’t really work, so don’t waste your money.”
The practical truth is somewhere in between. The advanced video analytics that are an important component of every CheckVideo solution are just the right amount of this well-evolved technology to solve 90% or more of the problems that have you thinking about achieving better security with video in the first place.
The advanced self-adapting video analytics software from CheckVideo enables the end-user to utilize video to proactively address security challenges, instead of waiting for something bad to happen.
The Power of Advanced Video Analytics
CheckVideo is the only video analytics security camera and video verification security system that has been hardened with more than 20 years of field-proven experience. CheckVideo solutions protect airports, critical infrastructure, outdoor assets and provide business intelligence to thousands of customers every day. The analytics technology in each solution goes well beyond motion detection and rules-based approaches that require significant training, and often suffer from high percentages of false alarms. CheckVideo analytics are calibration-free, automatically learn and adapt to the environment and require no training or manual adjustment. With the fastest learning rate in the industry, CheckVideo enabled cameras start detecting accurately within a few seconds. The analytics track and classify all moving objects, and include behavior recognition to only send alerts when certain types of activities occur. Best of all, this is available in a cost-effective package that can be managed and configured entirely off-site.
Video Analysis Process
Step 1 Create a Background Image
Step 2 Background Subtraction
Segment foreground moving pixels from backgrounds
Step 3 Clutter Removal
Remove all non-object pixels
Step 4 Tracking
A current object is matched with the best-fit object maintained in the track history
Step 5 Classification
Classified a labeled object into human(s), vehicle(s), or other objects based on machine learning
Step 6 Behavior Recognition
Based on analysis of a snapshot or a track history, report the object’s behavior with an alert
- Motion Detection: Detects moving objects in the scene, removes clutter to reduce false alarms.
- Zone Violation: Detects an object in a zone. Zones can be any polygon shape and unlimited
zones per view are supported.
- Camera Tamper: Detects scene changes, dark conditions, view blocked.
- Video Loss: Detects video signal loss.
- Person Detection: Qualifies a moving object as a person based on machine learning.
- Vehicle Detection: Qualifies a moving object as a vehicle based on machine learning.
- Loitering: Looks for people that stay in a zone for some period of time.
- Vehicle Stopped: Looks for vehicles that stay in a zone for some period of time.
- Additional video verification security analytics available upon request.
- Counting: Counts objects or people entering a zone or in a zone. Aggregates counts over
hours, days or weeks.
- Heat Maps: Provides motion, person or vehicle heat maps that show activity overlaid on the
view, aggregated over time.
- Dwell Time: Measures the average amount of time spent by a person or vehicle in a zone.
- Tracks: Overlays tracking information for people or vehicles on the view.
- Transaction: Correlates video with transaction events.