Trackers/Histogram

The Histogram Operator is a powerful real-time tool for analyzing video content based on content colors (hue, saturation and value, or HSV in short).

The core of the operator has one major function: count the number of instances of each unique color. This calculation can also be described as a HSV-histogram: A histogram is a mathematical representation of how common a certain value is. In this case, the histograms keeps track of the most common color = hue, saturation and value (HSV).

The operator is very flexible and can be used in many different ways and in many use-cases. This flexibility comes with a bit of a drawback: complexity. This operator is one of the most complex operators to describe and to understand.

The best way of thinking of this operator is as if it were a "filter": you define an HSV selection and additional criteria, and if those are fulfilled the operator will consider your content (video frame) to be qualified. If your criterias also include a Temporal filter (content over time), the operator will analyze the content over a number of frames and provide additional options for the content to be qualified.

The video analysis engine of the operator can be restricted to a certain part of the video (Processing Area).

The configuration below will detect a green area and show the qualified pixels as pink (To see the qualified pink pixels, uncheck the “Video pass-through” checkbox). Unchecking the “Show qualified pixels” will still detect the green pixels but not visually show them as pink.

• Configuration
  In this section, you define the criteria, or "filters", for qualifying pixels.

  • Minimum hue value - the minimum hue (0-360)

  • Maximum hue value - the maximum hue (0-360)

  • Minimum saturation - the minimum saturation (0-100 %)

  • Maximin saturation - the maximum saturation (0-100 %)

  • Maximum value - the maximum value (0 - 100%)

  • Minimum detection percentage - defines the minimum percentage of pixels that needs to qualify.

• Visualization

  • Show qualifying pixels - when selected, all qualifying pixels will be shown as pink pixels.

  • Hide disqualifying pixels - when selected, all pixels that don’t qualify will be hidden.

  • Hide all output - hides all pixels that are being processed.

• Temporal filter (time)

  • Consecutive Qualify Frames - the number of consequent frames that needs to qualify for the Temporal Qualified check to pass (become true).

  • Consecutive Un-qualify Frames - the number of consequent frames that needs to be disqualified for the Temporal Qualified check to pass (become true).

• Results

  • Qualified - true/checked if the frame qualified.

  • Temporal Qualified - true/checked if the frame is qualified using the temporal filter.

  • Most common hue - the most common hue in the frame

  • Count - the number of pixels having the most common hue

  • Number of hue values - the number of unique hue values

  • Most common saturation - the most common saturation in the frame

  • Number of saturation values - the number of unique saturation values

• Summary

  • Total number of pixels analyzed

  • Total number of pixels processed

  • Percent processed

• Processing area

  • Video-pass-through - when selected, the operator will not make any visual changes to the image.

  • Left - crop left

  • Right - crop right

  • Top - crop top

  • Bottom - crop bottom

Tutorial

One example on how to use the Histogram operator can be found in the Roulette Replay tutorial.