Configuring areas for object detection scenarios - KiwiVision™ 4.1 GA

KiwiVision™ Video Analytics Guide 4.1 GA

series
KiwiVision™ 4.1 GA
revised_modified
2019-07-10

You can define the areas from which you want detected changes to trigger events.

What you should know

You can enable the following options to visualize the sensitivity of your configuration:
Show scene background
Detected differences from the learned background are highlighted in blue if they are minor, in green if they are moderate, or in red if they are severe.
Show detection grid
Displays the analysis blocks that make up the defined area. A block outlined in red indicates the block is active.
NOTE: Depending on which type of scenario you are configuring, some settings might be unavailable or hidden if Advanced mode is disabled. When Advanced mode is enabled, the advanced settings are displayed in bold.

Procedure

  1. From the Config Tool home page, open the Video task.
  2. From the area view, select a camera that you applied an analytics scenario to.
  3. Click the Video analytics tab, and then click the scenario you want to configure.
  4. Click the Areas tab.
  5. Define the area as required:
    1. Click and drag the area to move it to the location you want.
    2. Click and drag the white dots to adjust the shape and size of the area.
    3. (Optional) Add or remove areas.
      • Add areas ()
      • Remove the selected area or the selected point ()
        NOTE: When an area or point is selected, it is yellow.
  6. In the Area name field, rename the selected area.
  7. In the Detection section, configure the following settings:
    Foreground detection tolerance
    The sensitivity of change detection. If set too high, differences from the background you want detected might not get detected. If set too low, the slightest changes are detected and might cause false alerts.
    Min. foreground per block
    The portion of a block that must differ from the learned background enough for the block to be considered foreground.
    Detection time
    The duration that a block must be detected as foreground to be considered active. The Detection time countdown begins when the specified Stationary object time period ends. This setting is used to minimize false alerts caused by rapid changes in a block, such as moving shadows.

    When Show detection grid is enabled, Detection time is represented by the bottom of the blocks where an object is detected turning red.

    Activity cooldown
    The duration that a block remains active after it is no longer detected as foreground.

    When Show detection grid is enabled, Activity cooldown is represented by the blocks outlined in grey. The bottom of the blocks turn red.

  8. In the Object size section, configure the following settings:
    Min. blocks per object
    The minimum number of active blocks required for an alert to be triggered.
    Max. blocks per object
    The maximum number of active blocks required for an alert to be triggered.
    Detect all active blocks
    When this option is selected, active blocks anywhere in the defined area can trigger alerts.
    Detect only connected blocks
    When this option is selected, only active blocks that are adjacent to each other can trigger alerts. This option minimizes false alerts caused by scattered movement, such as camera noise.
  9. Click Apply.

Example

The following examples show the stages leading up to an alert. In these examples, the motion analysis mode is set to Detect stationary objects.

Detection time

The following image shows that an object is detected, and the Detection time is counting down. This countdown is indicated by the bottom of the analysis blocks turning red from left to right.

Active blocks

The following image shows that the blocks are active; they are outlined in red.