+
    ÿäûjH  ã                   óô   € R t ^ RIHt ^RIHtHtHtHt ^RIHtH	t	 ^RI
H
t
 ^RIHtHt ^RIHtHt ^RIHtHtHtHtHt ^RI5 ^R	IHtHt ^R
IHt ^RIHt ^RIHt ^RIH t H!t! ^RI"H#t#H$t$H%t%H&t& . ROt'R# )a}  Contributed image operations and task-level model wrappers.

Which function or class to use, with the module form after the function:

- Cut an image into sliding-window patches: :func:`extract_tensor_patches`, :class:`ExtractTensorPatches`; put them
  back, averaging overlaps: :func:`combine_tensor_patches`, :class:`CombineTensorPatches`; the padding that makes the
  windows fit: :func:`compute_padding`.
- Label 8-connected foreground components: by a fixed number of pooling steps, :func:`connected_components`;
  exactly, with no step budget, :func:`connected_components_union_find`.
- Euclidean distance to the nearest positive pixel, in 2D or 3D: :func:`distance_transform`,
  :class:`DistanceTransform`.
- Match the histogram of an image to a template: :func:`histogram_matching`; the 1-D linear interpolation it uses,
  which extrapolates the end segments: :func:`interp`.
- Plasma-fractal noise maps: :func:`diamond_square`.
- Cluster points by k-means: :class:`KMeans`.
- Run a Python callable as a module: :class:`Lambda`.
- Audit the geometry of one augmentation pipeline call: :func:`audit`, which returns an
  :class:`AugmentationAuditReport` of :class:`AugmentationAuditStep` and :class:`SpatialAudit` entries.
- Detect edges with a wrapped network: :class:`EdgeDetector`; build one around DexiNed, pretrained by default:
  :class:`EdgeDetectorBuilder`.
- Detect faces and their landmarks: :class:`FaceDetector`.
- Detect objects: :class:`ObjectDetector`, built by :class:`RTDETRDetectorBuilder`.
- Upscale images: :class:`SuperResolution`, configured by :class:`SuperResolutionConfig` and built by
  :class:`SmallSRBuilder` or :class:`RRDBNetBuilder`.
- Stitch images with overlapping fields of view: :class:`ImageStitcher`.
- Segment one image from point, box and mask prompts: :class:`~kornia.contrib.visual_prompter.VisualPrompter`.
- Classify images with the TinyViT network: :class:`TinyViT`.
)ÚTinyViT)ÚAugmentationAuditReportÚAugmentationAuditStepÚSpatialAuditÚaudit)Úconnected_componentsÚconnected_components_union_find)Údiamond_square)ÚDistanceTransformÚdistance_transform)ÚEdgeDetectorÚEdgeDetectorBuilder)ÚCombineTensorPatchesÚExtractTensorPatchesÚcombine_tensor_patchesÚcompute_paddingÚextract_tensor_patches)Ú*)Úhistogram_matchingÚinterp)ÚImageStitcher)ÚKMeans)ÚLambda)ÚObjectDetectorÚRTDETRDetectorBuilder)ÚRRDBNetBuilderÚSmallSRBuilderÚSuperResolutionÚSuperResolutionConfigN)r   r   r   r
   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r   r	   r   r   r   r   )(Ú__doc__Úkornia.models.tiny_vitr   Úaugmentation_auditr   r   r   r   r   r   r	   r   r
   Úedge_detectionr   r   Úextract_patchesr   r   r   r   r   Úface_detectionr   r   Úimage_stitchingr   Úkmeansr   Úlambda_moduler   Úobject_detectionr   r   Úsuper_resolutionr   r   r   r   Ú__all__© ó    ÚSC:\Ai_Coding\apps\studio\ComfyUI\.venv\Lib\site-packages\kornia\contrib\__init__.pyÚ<module>r.      sL   ðñ$õ: +ç cÓ cß WÝ *ß Eß =÷õ ô ß :Ý *Ý Ý !ß Cß dÓ dò‚r,   