Skip to main content

Most dense color threshold detection

most_dense_color_threshold_detection applies the same HSV thresholding as color_threshold_detection, then returns only the candidate with the largest or smallest contour area.

Inputs​

  • frame: a three-channel BGR NumPy image.

Outputs​

  • detection: a list containing at most one detection dictionary with bbox, class_id, color_name, and area.

When to use​

Use this operation when color thresholding may find several regions but the pipeline needs only the largest or smallest one. Use color_threshold_detection when every matching region matters.

Configuration​

It supports the same camera_bus_id, target_size, color_ranges, area limits, blur, and morphology settings as Color threshold detection.

  • selection_mode: most_dense selects the largest contour area. least_dense selects the smallest. Default most_dense.

Important behavior and limitations​

Density means contour area, not filled-pixel ratio or physical density. The result can change when lighting, thresholds, morphology, or camera perspective changes. An empty list means no candidate passed the configured filters.