typed op• 数据种类:qimage → rgbimage
• 调用:fullseye.apply(img, "tb_quaternion_to_rgb", a=0.5, b=0.5)(2-D 的模型是一张图 + 两个标量旋钮 a,b∈[0,1])
*无图: 该算子以 qimage 为输入。从图像开始的 Studio 程序无法到达该类型，请参见下方的可运行示例。*
四元数图像的向量部分，作为线性 RGB。→ (H, W, 3)。
> 以下的详细说明为原文 —— 摘要与标题已翻译。
The inverse of :func:rgb_to_quaternion — and, by default, a *checked*
inverse. A quaternion image that picked up a scalar component somewhere (a
Hamilton product with a non-pure quaternion, a monogenic signal handed here
by mistake) is refused rather than silently truncated, because dropping
the `w` component is exactly the kind of loss that produces a plausible
picture from the wrong data. Pass `allow_scalar=True` to opt in to the
truncation when it is what you meant.
The tolerance is relative to the field's own peak modulus
(:data:_MONOGENIC_K_TOL, 1e-9): a quaternion image that really is pure
carries `|w|` at the 1e-17 level after a round trip through two FFTs, and
anything with a meaningful scalar part is many orders above that. Nothing
real lives in between.
Raises `ValueError: *qimage* is not a finite (H, W, 4)` array; or
it has a non-negligible scalar part and `allow_scalar` is False.
Typed bridge of the quat op `quaternion_to_rgb into the 2-D evolution registry: the same implementation, called under the op(v, a, b) convention. This op has no tunable parameter; a and b` are unused.
• 示例数据目录(下载 URL / 许可证) —— 2-D 用 skimage.data(BSD/公有领域)加合成图,3-D 给出真实数据源(Stanford/PDS 等)的下载 URL。
• 算子来历与参考文献 —— 该算子族所依据的研究/方法出处。
• (暂无)
rgbimage 作为输入)identity · tb_wetness · tb_sensor_capture · tb_specular_diffuse_split · tb_specular_coefficient_map · tb_specular_free_transform · tb_rgb_to_quaternion
typed)tb_points_to_voxel · tb_estimate_point_normals · tb_iss_keypoints · tb_angle_3points · tb_project_points · tb_render_point_depth · tb_statistical_outlier_removal · tb_radius_outlier_removal
*Provenance: ops.py — 2D 算子登记表。本条目由 tools/opdocs.py md 自动生成(请勿手工编辑)。*
© 2026 Kazufumi Furuse — Fullseye operator documentation. Licensed under Apache-2.0.