features op• Data kinds: image → feature
• Call: fullseye.apply(img, "cv_hough_circles", a=0.5, b=0.5) (the 2-D model is one image plus two scalar knobs a,b∈[0,1])
• HALCON equivalent: hough_circles (the HALCON reference is a useful guide to its meaning and parameters)
• GPU: none (CPU reference implementation only)
*The figure is the actual output on a synthetic 128×128 input. Left: input, right: output. Point clouds are drawn as a top-down scatter (brightness = z), 1-D series as a line plot, volumes as the maximum-intensity projection along z, videos as the middle frame, complex images as magnitude; return values that are not pictures are shown as the values themselves.*
*Knob a does not change the output (measured: identical at 0.1 / 0.5 / 0.9).*
*Knob b does not change the output (measured: identical at 0.1 / 0.5 / 0.9).*
Circle detection via the Hough transform (a single scalar feature, OpenCV implementation). Detects circle centers and radii by voting, using the HOUGH_GRADIENT method that relies on edge gradient information - here, only the number of detected circles is returned (0 if none).
Corresponds to HALCON's hough_circles (Detect centers of circles for a specific radius using the Hough transform.) (approximate: returns only the count, not the centre coordinates). Implemented as cv2.HoughCircles(_u8(v), HOUGH_GRADIENT, dp=1, minDist=10+int(a*20), param1=100, param2=20+int(b*20), minRadius=3, maxRadius=20) — a sets the minimum distance between detected centres to 10–30, b sets the centre-voting threshold param2 (smaller gives more false detections) to 20–40. param1 (the internal Canny high threshold) and the radius range (3–20) are fixed.
Precondition: smooth first (this op does not smooth internally). HOUGH_GRADIENT votes for centres along each pixel's gradient direction, so noise that scatters the directions keeps the votes from accumulating. OpenCV's description and examples also blur the input before passing it (GaussianBlur / medianBlur). Measured (2026-11-03, 96x96, a 0.6 disc on a 0.3 background, noise σ 0.02): without smoothing the count is 0 both for three discs of radius 6–8 and for one disc of radius 18 (calling cv2.HoughCircles directly also returns None). Applying gaussian a=0.5 (σ about 1.65 px) first counts 3 / 4 correctly. A result of 0 cannot tell "no circle" from "the votes were split", so on noisy images always put smoothing in front.
• gallery2d_features family guide
• Sample-data catalog (download URLs / licences) — 2-D uses skimage.data (BSD/public domain) plus synthetic images; 3-D lists download URLs for real data sources (Stanford, PDS, …).
• Operator provenance and references — the sources of the research/methods this op family came from.
• The canonical algorithm (author, year) and its uses are named in the family usage guide above.
The program below has been verified to run (same input as the figure). In Studio's help this block becomes buttons that load and run it on the spot.
cv_hough_circles 0.50 0.50
▸ Load this pipeline · Load & run
• gallery2d_features — py -3.11 examples/gallery2d_features.py
feature as input)features)effective_bit_depth · blob_count · area_frac · count_contours · total_length · vol_count · sk_euler · sk_entropy_feat
*Provenance: ops.py — 2D operator registry. This per-op note is generated by tools/opdocs.py md (do not hand-edit).*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.