| Customize Help

Basic concepts for the MIL Color Analysis module



The basic concepts and vocabulary conventions for the MIL Color Analysis module are:

  • Area identifier image. An image, provided with the color matching operation, that specifies the target areas.

  • Background pixels. Pixels outside the target areas.

  • Best-matched color-sample. The color-sample that most matches a target area's color, with respect to all color matching constraints.

  • Color distance. The numerical difference between two colors.

  • Color-sample element. The individual color data entities that define a color-sample.

  • Color identification. Matching the color of each target area with the best color-sample.

  • Color matching. Calling McolMatch() to perform color identification or supervised color segmentation.

  • Color matching context. A MIL object that stores all color-samples and color matching settings.

  • Color-sample. The information, in the context, that defines the color to be processed using relative color calibration or color matching. For relative color calibration, color-samples have an associated color-mapping between it and the reference color-sample.

  • Color space. A mathematical model, that typically has 3 components (for example, RGB, HSL, LAB), with which to describe colors.

  • Color space encoding. Color transformation from the range represented in an image buffer (for example, 0 to 255) to its native (theoretical) range (for example, all real numbers between 0 and 1).

  • First principal component. The strongest principal component vector computed from a PCA. If no principal component is the strongest, one is arbitrarily considered the first principal component.

  • IEC. Refers to the International Electrotechnical Commission, a globally-recognized standards organization in the field of electrotechnology.

  • Match score. A measure of similarity between the color of the target area and the color of the color-sample, when performing color matching.

  • Outliers. Pixels, within a target area, that do not relate to any color-sample.

  • Principal component analysis (PCA). A mathematical analysis which, when applied to a color image's data, results in vectors pointing towards the direction of maximum color variance. Each vector is considered a principal component. There are as many principal components as there are color bands.

  • Reference color-sample. The information, in the relative color calibration context, with which to establish the color mapping used in the relative color calibration.

  • Reference color space. The standard used to interpret the color space data (for example, sRGB).

  • Relative color calibration. Calling McolTransform() to adjust an image's color data according to the mapping associated with a color-sample in a relative color calibration context.

  • Relative color calibration context. A MIL object that stores the reference color-sample, the color-samples, and the color mapping with which to perform relative color calibration.

  • Relevance score. A measure of confidence associated with the color match score.

  • Source color space. The color space with which the MIL Color Analysis module interprets color.

  • sRGB. Standard RGB specifications, as defined by the IEC Project Team 61966-2-1.

  • Supervised color segmentation. Matching the color of each target area pixel with the best color-sample. Used to determine the proportion of color in a target area, based on the color-samples.

  • Target. The image with which to perform the color matching.

  • Target area. A section of the target with which to match a color-sample.

  • Triplet. A color-sample defined with three explicit color component values.