diabolo op• Data kinds: rgb × table × table → table
• Call: import fullseye as fs; fs.ledger.diabolo_axis_from_image(img, cam: 'dict', params: 'dict', *, refine: 'bool' = True, iters: 'int' = 15) -> 'dict' (to call the implementation directly, import diabolo; diabolo.diabolo_axis_from_image(img, cam: 'dict', params: 'dict', *, refine: 'bool' = True, iters: 'int' = 15) -> 'dict'; from the registry, opsdrive.get("diabolo_axis_from_image"))
• GPU: none (CPU reference implementation only)
From 1 frame, finds the center of the near cup's rim center_rim and the axis axis (pointing toward the camera). Diabolo center center = center_rim − (L/2) axis. No learning.
Initial value (weak perspective): semi-major axis of edge ellipse → depth, short/long = cos θ, shift of center of bottom plate = (L/2 − z_h) sin θ, θ = atan2(sin, cos) (avoid acos floor), direction of deviation = direction of image of axis. Finishing: Edge circle (radius R, center C) and bottom plate (radius r_h, center C − (L/2 − z_h) a) Match the polygonal moments (area, center of gravity, quadratic) of exact perspective projection to the moments of the mask using Gauss–Newton (5 degrees of freedom, 8 residuals). ok = False if the residual residual_px exceeds 0.15 px (e.g. the bottom plate is hidden by a wall, 2.8 px at 45° from line of sight, ≤ 0.02 px up to 38°).
Returns {"center", "center_rim", "axis", "axis_init", "tilt_init_deg", "a_px", "b_px", "hub_px", "rim_px", "residual_px", "ok", "seg"}. Raises ValueError: The shape of the image, the inside of the cup or the bottom plate cannot be seen (cannot be estimated).
• 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.
• poc_diabolo_model_and_vision — py -3.11 examples/poc_diabolo_model_and_vision.py
table as input)course_layout · course_occupancy · course_contains · world_build · world_camera · world_move · lidar_scan · rss_longitudinal_same
diabolo)diabolo_params · diabolo_spheroid · spheroid_closest · diabolo_dynamics_step · diabolo_simulate · diabolo_state_sequence · diabolo_throw_catch_truth · string_tension_static
*Provenance: diabolo.py — DRIVE 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.