Color Detection
Overview
Color Detection node finds regions that match configured color rules and returns their outlines as detection contours.
Use this node to locate objects by color in scenes where the target has a consistent, distinguishable hue — for example colored caps on bottles, zone markings on a floor, or colored indicators on equipment. Pair it with downstream nodes such as Detection Count Zones or logic nodes that act on the contour data.
Input
Input Image
image requiredThe image frame to analyze. Connect this to a camera or upstream image output.
Color Label Thresholds
array requiredColor rules used to isolate target-colored regions. Each entry includes a color threshold, optional label, and overlay color. Duplicate labels are merged using the first overlay color; empty labels become "color". See Color Thresholds for threshold details.
Morphology
object requiredMorphological operations to clean the color mask before contour extraction. See Morphology for details.
Top contours keep amount
integer advancedKeep the top number of contours, sorted by area from largest to smallest.
Range: 0 to 10000
Default: 500
Contour Width Range
array advancedContour filter using width as [min, max] pixels.
Default: [5, 1000]
Contour Height Range
array advancedContour filter using height as [min, max] pixels.
Default: [5, 1000]
Overlay results
boolean required advancedWhether to draw detection contours and filled regions on the output frame. See Overlay Results.
Default: true
Overlay opacity
number required advancedOpacity of the colored region overlays. Shown when Overlay results is enabled.
Range: 0.0 to 1.0
Default: 0.5
Draw bounding boxes
boolean required advancedWhether to draw bounding boxes on the image frame. Shown when Overlay results is enabled.
Default: true
Output
Overlay Image
imageOutput frame from the node. If overlays are enabled, detected color regions are filled and outlined.
Detected Count
integerNumber of color-matched regions detected in the current frame.
Detected Objects
arrayArray of region detection objects. Each object contains:
- Segmented Contour Points (
contour) - Bounding Box (
bbox) —[x, y, width, height] - Label (
label)
Binary Mask Image
imageBinary mask produced from the color threshold step. Useful for debugging threshold rules — white pixels are the regions that passed the color match.