Video Analytics

Video Analytics Brochure

Video Analytics Brochure

6.42 MB

The CathexisVision Video Analytics Suite provides a powerful enhancement to the world-class, feature-rich CathexisVision Video Surveillance Management software, and signicantly enhances the value of any CCTV surveillance installation. CathexisVision goes beyond surveillance to offer the most sophisticated and fully integrated infrastructure and security management system that delivers maximum return on surveillance investment.

Seamless integration, easy installation and exceptional ease-of-use have made CathexisVision the preferred surveillance software solution for the global retail, banking, mining, education, residential, commercial, healthcare, transport, logistics and city surveillance sectors.

The CathexisVision Video Analytics Suite includes:

  • Presence Detection
  • Enter/Exit Area
  • Line-Crossing
  • Motion Detection
  • Time or Stop in Area
  • Direction
  • Speed
  • Loitering
  • Object Detection
  • Appear/Disappear
  • Object Counting
  • Queue Length Analysis
  • License Plate Recognition

Algorithm Set-Up Features

Our experience over the past 20 years has shown us that Video Analytics require an empirical approach to configuration, and that more sophisticated present complex and challenging configuration. The Cathexis team has therefore developed the ability to configure the analytics on pre-recorded and live video footage. This enables users to adjust the algorithm until the user is completely comfortable with its performance. In addition, Cathexis provides feedback on the performance of the algorithm in real-time, making the set-up of the algorithms easier and more intuitive.

Within its various algorithms, CathexisVision has provide the flexibility, functionality and ease-of use to provide the most accurate results. These include:

  • Inclusion zones
  • Size filters
  • Calibration
  • Image stabilisation
  • Shadow suppression
  • Direction settings

Analytics algorithms and applications

Presence Trigger

The user has the ability to draw a multi-point inclusion zone, which will trigger an event on the presence of an object in the specic area. As a result, any object that is detected according to the set-up parameters, and is detected in this zone, will generate a trigger.

Application Example:

  • Movement in the passageway of a building will trigger recordings and an alarm.

Enter and exit area

This feature forms part of the Cathexis Advanced Area Analytics and the algorithm allows for the configuration of object entry or exit detection into or out of a selected area or inclusion zone. The area for detection is selected via a multi-point configuration tool, and the algorithm can be set to trigger on objects entering the area, leaving the area, or both.

Application Example:

  • Notification when a vehicle enters a restricted area of a carpark.
  • A person entering an area in an oil refinery or mining facility where workers are already active.

Stopping in area

Similar to the Loitering Algorithm, this is more specific in that the trigger will only be activated if an object has entered and then “stopped” in the inclusion zone for longer than a pre-determined time.

Application Example:

  • A vehicle that has stopped on the shoulder lane of a freeway.
  • A person entering a waiting area of a bank and waiting for longer than a specified time.

Line cross

An imaginary line is placed in a specic place on the video image, and any objects of pre-determined size that cross this line will generate a trigger. The algorithm can be configured to detect forward, reverse or both directions across the line.

Application Example:

  • Customers entering and exiting the entrance of a retail store.
  • Intruders crossing a boundary line or fence line.
  • Pedestrians walking across a highway

Advanced Line cross

This algorithm provides an extension to the basic algorithm by firstly, providing an articulated line with as many points as required (Example: if you want a curved line) and, secondly, by provide a double line that will then only trigger if both lines have been crossed.

Application Example:

  • An intruder entering a “no-mans-land” area will be tracked and detected but only once they have crossed both lines will the alarm be triggered.
  • An articulated line around a perimeter.

Loitering

This algorithm allows for the configuration of objects in a certain area based on a set timer. An alarm will be triggered if the object has been present in a specic area that exceeds the pre-determined time.

Application Example:

  • An alarm can be sounded if a person is loitering near an ATM machine for longer than a specified period of time.

No Motion

The No Motion option is normally used in a control room environment to monitor the control room operators. If there is no movement detected in the operator seating area of the control room for a specied period of time an alarm will be triggered.

Application Example:

  • The No Motion algorithm could typically be used in a control room environment to monitor the control room operators. If No Motion has been detected in the operator seating area for 30 minutes, an alarm will be triggered. No Motion detected may indicate that the operator has fallen asleep or left the control room.

Speed Measurement

This algorithm enables the measurement of the speed of a tracked object across a pre-determined distance. The speed triggers can be configured in either a forward or reverse direction.

Application Example:

  • The speed between two points that a vehicle is travelling in a certain direction
  • A person running instead of walking.

Static Object Detection

Object Detection is used to monitor and trigger user-defined actions based on stationary objects. If an object is detected in a pre-dened area for longer than a specied time, the system will generate a trigger.

Application Example:

  • A suitcase left in an airport public area.
  • An illegally parked vehicle or an abandoned parcel in a public area.
  • An object left blocking a fire escape area.

Queue Length Analysis

The Queue Length Algorithm is used to detect if a queue of people or objects is occupied for too long. This is done by creating multiple zones along the queue area and then generating a trigger if the zones within the queue are occupied for longer than a specied time.

Application Example:

  • The queue for the tills at a retail store, so that the store manager can be notified if customers are not being served quickly enough.
  • Queue at a carpark.

Automatic Number Plate Recognition (ANPR)

Also known as LPR (License Plate Recognition), we make use of a Character Recognition Algorithm through which the license plate details of vehicles can be identied. The sophisticated algorithm has been proven in more than 200 countries and works with both standard visible light cameras and specialised ANPR cameras, depending on the requirements and the environment.

Application Example:

  • Queue at a carpark.Identifying stolen vehicles in cities.
  • Access control for individuals or groups for Hotels or Carparks.
Share with
Share on FacebookShare on LinkedInTweet about this on TwitterShare on Google+Email this to someone