__header__

Description

Finds the the primary colors, in an image using the Signature space.
The output is a 8bit image with a color palette describing the colors at the specific greylevels
typically red =60
green = 120
blue = 180

Inputs

CIFViewer

CIF

Color1
Color2
Color3

These are optional, and define the target train colors.
By default the system defines Red Green Blue
as red 1,0.7,0.7
as green 0.7,1.0,0.7
as blue 0.7,0.7,1.0
These act as the train color for the tool

Parameters

CIFLUT

output is the LUT and this could be passed to another tool like Blob

minWhite1
minWhite2
minWhite3

defines the threshold for white
colro in this range are not set to white

maxBlackX1
maxBlackX2
maxBlackX3

defines the threshold for black Colors in this range are set to black

SignatureThreshold

define the minimum separation for bewteen colors Color not haveing this separation are set to black

Bias

The Bias - uses in the Filter Distance method

BitDepth

the bit depth of the LUT

DistanceMethod

the Filter distance method

of interest are
IPCOLORDIFFGeneric:
Sum Off Positive Component Distance from the Trained Color

IPCOLORDIFFGenericSQR:
Sum off Positive Squared Distance from the Trained Color

IPCOLORDIFFGenericSQRT:
Sum of Componet Positive Square root Distance from the Trained Color

DoSpatialSize
SpatialFilterSize

if set we eliminate isolated rich colors

UpdateViewer

Update the image to the viewer

RefreshViewer

Refresh the image

Image

The output image
__footer__