qiUdZddlmZddlZddlmZmZmZddl m Z m Z ddl m Z mZmZmZmZmZmZmZmZmZmZmZmZmZddlmZdd lmZmZm Z dd l!m"Z"erddl#Z#erd d l$m%Z%er ddl&Z&dd l&m'Z'm(Z(ejRe*Z+ ddejXde-e zdzfdZ. ddede-e zdzdefdZ/defdZ0GddedZ1e"dGddeZ2dgZ3y)z%Image processor class for SuperPoint.) TYPE_CHECKINGN)BaseImageProcessor BatchFeature get_size_dict)resizeto_channel_dimension_format)ChannelDimension ImageInput ImageTypePILImageResamplingget_image_typeinfer_channel_dimension_format is_pil_imageis_scaled_imageis_torch_availableis_valid_imageis_vision_availableto_numpy_array valid_imagesvalidate_preprocess_arguments) ImagesKwargs) TensorTypeloggingrequires_backends)requires)SuperGlueKeypointMatchingOutput)Image ImageDrawimageinput_data_formatc|tjk(rQ|jddk(rytj|d|dk(xrtj|d|dk(S|tj k(rQ|jddk(rytj|d|d k(xrtj|d |d k(Sy) NrrTr.r...r.r.r')r FIRSTshapenpallLAST)r!r"s j/opt/pipecat/venv/lib/python3.12/site-packages/transformers/models/superglue/image_processing_superglue.py is_grayscaler29s,222 ;;q>Q vveFmuV}45`"&&vRWX^R_A_:`` .33 3 ;;r?a vveFmuV}45`"&&vRWX^R_A_:`` 4returncttdgt|tjrt ||r|S|t jk(r7|ddz|ddzz|ddzz}tj|gd zd }|S|t jk(r5|d dz|d dzz|ddzz}tj|gd zd }St|tjjs|S|jd}|S)a4 Converts an image to grayscale format using the NTSC formula. Only support numpy and PIL Image. This function is supposed to return a 1-channel image, but it returns a 3-channel image with the same value in each channel, because of an issue that is discussed in : https://github.com/huggingface/transformers/pull/25786#issuecomment-1730176446 Args: image (Image): The image to convert. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format for the input image. visionr"r$gŏ1w-!?r%gbX9?r&gv/?rr)axisr)r*r+r(L) rconvert_to_grayscale isinstancer.ndarrayr2r r,stackr0PILrconvert)r!r" gray_images r1r:r:Hs "*XJ7%$ 1B CL  0 6 6 6v/%-&2HH5QW=[aKaaJ:,"2;J"2"7"7 7v/%-&2HH5QW=[aKaaJ:,"2lenr-)r!s r1_is_valid_imagez8validate_and_format_image_pairs.._is_valid_imagevsGE" 5 ! fnU&;y}}&L fQTUZU`U`QaefQf r3r'c3.K|] }|ywN.0r!rEs r1 z2validate_and_format_image_pairs..}s#Q_U%;#Qc3K|]:}t|txr$t|dk(xrtfd|D<yw)r'c3.K|] }|ywrGrHrIs r1rKz..sCuOE*CrLN)r;listrDr/)rJ image_pairrEs r1rKz2validate_and_format_image_pairs..sN  z4 ( DJ1$ DC CC D sAA)r;rOrDr/ ValueError)rA error_messagerPr!rEs @r1validate_and_format_image_pairsrSmsM &$ v;! #Q&#Q QM  %   -3Kj KuEKEK K ] ##LsA3ceZdZUdZeed<y)SuperGlueImageProcessorKwargsz do_grayscale (`bool`, *optional*, defaults to `True`): Whether to convert the image to grayscale. Can be overridden by `do_grayscale` in the `preprocess` method. do_grayscaleN)__name__ __module__ __qualname____doc__bool__annotations__rHr3r1rUrUs r3rUF)total)torch)backendsceZdZdZdgZddej dddfdedee e fdzded ed e d ed dffd Z dde jdee e fde ezdzde ezdzfdZdddddddej"df dedzdee e fdzdedzd edzd e dzd edzde ezdzdede ezdzd efdZ ddddeeezde d eee ej0ffdZdedeee ej0fd edfdZdZxZS) SuperGlueImageProcessorap Constructs a SuperGlue image processor. Args: do_resize (`bool`, *optional*, defaults to `True`): Controls whether to resize the image's (height, width) dimensions to the specified `size`. Can be overridden by `do_resize` in the `preprocess` method. size (`dict[str, int]` *optional*, defaults to `{"height": 480, "width": 640}`): Resolution of the output image after `resize` is applied. Only has an effect if `do_resize` is set to `True`. Can be overridden by `size` in the `preprocess` method. resample (`PILImageResampling`, *optional*, defaults to `Resampling.BILINEAR`): Resampling filter to use if resizing the image. Can be overridden by `resample` in the `preprocess` method. do_rescale (`bool`, *optional*, defaults to `True`): Whether to rescale the image by the specified scale `rescale_factor`. Can be overridden by `do_rescale` in the `preprocess` method. rescale_factor (`int` or `float`, *optional*, defaults to `1/255`): Scale factor to use if rescaling the image. Can be overridden by `rescale_factor` in the `preprocess` method. do_grayscale (`bool`, *optional*, defaults to `True`): Whether to convert the image to grayscale. Can be overridden by `do_grayscale` in the `preprocess` method. pixel_valuesTNgp? do_resizesizeresample do_rescalerescale_factorrVr4c t|di|||nddd}t|d}||_||_||_||_||_||_y)Nii)heightwidthFdefault_to_squarerH) super__init__rrcrdrerfrgrV) selfrcrdrerfrgrVkwargs __class__s r1rnz SuperGlueImageProcessor.__init__s^ "6"'tc-JTU;"   $,(r3r! data_formatr"c Lt|d}t|f|d|df||d|S)aL Resize an image. Args: image (`np.ndarray`): Image to resize. size (`dict[str, int]`): Dictionary of the form `{"height": int, "width": int}`, specifying the size of the output image. data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format of the output image. If not provided, it will be inferred from the input image. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - `"none"` or `ChannelDimension.NONE`: image in (height, width) format. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format for the input image. If unset, the channel dimension format is inferred from the input image. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - `"none"` or `ChannelDimension.NONE`: image in (height, width) format. Frkrirj)rdrrr")rr)ror!rdrrr"rps r1rzSuperGlueImageProcessor.resizesE:TU;  x.$w-0#/     r3return_tensorsc  (||n |j}||n |j}||n |j}||n |j}||n |j}||n |j }t |d}t|}t|s tdt||||||D cgc] } t| }} t|dr|rtjd| t|d} g} |D]]} |r|j!| ||| } |r|j#| || } |r t%| | } t'| | | } | j)| _t+dt-| d Dcgc] }| ||d z }}d |i}t/|| Scc} wcc}w)ad Preprocess an image or batch of images. Args: images (`ImageInput`): Image pairs to preprocess. Expects either a list of 2 images or a list of list of 2 images list with pixel values ranging from 0 to 255. If passing in images with pixel values between 0 and 1, set `do_rescale=False`. do_resize (`bool`, *optional*, defaults to `self.do_resize`): Whether to resize the image. size (`dict[str, int]`, *optional*, defaults to `self.size`): Size of the output image after `resize` has been applied. If `size["shortest_edge"]` >= 384, the image is resized to `(size["shortest_edge"], size["shortest_edge"])`. Otherwise, the smaller edge of the image will be matched to `int(size["shortest_edge"]/ crop_pct)`, after which the image is cropped to `(size["shortest_edge"], size["shortest_edge"])`. Only has an effect if `do_resize` is set to `True`. resample (`PILImageResampling`, *optional*, defaults to `self.resample`): Resampling filter to use if resizing the image. This can be one of `PILImageResampling`, filters. Only has an effect if `do_resize` is set to `True`. do_rescale (`bool`, *optional*, defaults to `self.do_rescale`): Whether to rescale the image values between [0 - 1]. rescale_factor (`float`, *optional*, defaults to `self.rescale_factor`): Rescale factor to rescale the image by if `do_rescale` is set to `True`. do_grayscale (`bool`, *optional*, defaults to `self.do_grayscale`): Whether to convert the image to grayscale. return_tensors (`str` or `TensorType`, *optional*): The type of tensors to return. Can be one of: - Unset: Return a list of `np.ndarray`. - `TensorType.PYTORCH` or `'pt'`: Return a batch of type `torch.Tensor`. - `TensorType.NUMPY` or `'np'`: Return a batch of type `np.ndarray`. data_format (`ChannelDimension` or `str`, *optional*, defaults to `ChannelDimension.FIRST`): The channel dimension format for the output image. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - Unset: Use the channel dimension format of the input image. input_data_format (`ChannelDimension` or `str`, *optional*): The channel dimension format for the input image. If unset, the channel dimension format is inferred from the input image. Can be one of: - `"channels_first"` or `ChannelDimension.FIRST`: image in (num_channels, height, width) format. - `"channels_last"` or `ChannelDimension.LAST`: image in (height, width, num_channels) format. - `"none"` or `ChannelDimension.NONE`: image in (height, width) format. FrkzSInvalid image type. Must be of type PIL.Image.Image, numpy.ndarray, or torch.Tensor)rcrdrerfrgrzIt looks like you are trying to rescale already rescaled images. If the input images have pixel values between 0 and 1, set `do_rescale=False` to avoid rescaling them again.)r!rdrer")r!scaler"r7)input_channel_dimr'rb)data tensor_type)rcrerfrgrVrdrrSrrQrrrlogger warning_oncerrrescaler:r appendrangerDr)rorArcrdrerfrgrVrtrrr"rpr! all_imagesi image_pairsrxs r1 preprocessz"SuperGlueImageProcessor.preprocesssp"+!6IDNN '38#-#9Zt +9+E4K^K^'3'?|TEVEV 'tTYYTU;18F#rs s%!)  6<vay I   %E %dXar s 5Zk l,UFWX/{VghE   e $ %7BB?=6Ts #F (Foutputsr target_sizes thresholdc|jjdt|k7r tdt d|Ds tdt |t r,tj||jj}n1|jddk7s|jddk7r td|}|jj}||jdjddddz}|jtj}g}t!|j||j"d d df|j$d d dfD]v\}}} } |ddkD} |ddkD} |d| } |d| }| | }| | }||kD|dkDz||jdkz}| |}|||}||}|j'|||d x|S) a Converts the raw output of [`SuperGlueKeypointMatchingOutput`] into lists of keypoints, scores and descriptors with coordinates absolute to the original image sizes. Args: outputs ([`SuperGlueKeypointMatchingOutput`]): Raw outputs of the model. target_sizes (`torch.Tensor` or `list[tuple[tuple[int, int]]]`, *optional*): Tensor of shape `(batch_size, 2, 2)` or list of tuples of tuples (`tuple[int, int]`) containing the target size `(height, width)` of each image in the batch. This must be the original image size (before any processing). threshold (`float`, *optional*, defaults to 0.0): Threshold to filter out the matches with low scores. Returns: `list[Dict]`: A list of dictionaries, each dictionary containing the keypoints in the first and second image of the pair, the matching scores and the matching indices. rzRMake sure that you pass in as many target sizes as the batch dimension of the maskc38K|]}t|dk(yw)r'N)rD)rJ target_sizes r1rKzISuperGlueImageProcessor.post_process_keypoint_matching..rsI[3{#q(IszTEach element of target_sizes must contain the size (h, w) of each image of the batch)devicerr'r(N) keypoints0 keypoints1matching_scores)maskr-rDrQr/r;rOr^tensorr keypointscloneflipreshapetoint32zipmatchesrr})rorrrimage_pair_sizesrresults mask_pairkeypoints_pairrscoresmask0mask1rrmatches0scores0 valid_matchesmatched_keypoints0matched_keypoints1rs r1post_process_keypoint_matchingz6SuperGlueImageProcessor.post_process_keypoint_matchingZs, <<  a C $5 5qr rILIIst t lD )$||LATATU !!!$)\-?-?-Ba-G j , %%++-  0 5 5b 9 A A"aA NN LL- := LL)W__QT%:G>.1D)4\)B)I)I!)L &L,)4\)B)I)I!)L &L,VYlL, TeHfW R [+{N7  +{V/C[Q  kAo{Q aQ\_`Q`aho p  6)A-{Q f@TWX@XZehiZij   NN> *7 +8A=Ks G G%cNtdd|z z}td|z}d}|||fS)zMaps a score to a color.rr)int)roscorergbs r1rz"SuperGlueImageProcessor._get_colors4 q5y! " e  1ayr3)NN)g)rWrXrYrZmodel_input_namesr BILINEARr[dictstrrfloatrnr.r<r rr,rrrrOtupler^Tensorrr rr __classcell__)rqs@r1rarasD,((&*'9'B'B '!))38nt#)% )  )  )) )46:;? % zz% 38n% ++d2 % !11D8 % T"&&*.2"&'+$(26(8(>(>;?oC$;oC38nt# oC %t+ oC 4K oC oCTkoCj(4/oC&oC!11D8oC oCj B2B!4;.B B d3 $% & BJ44#'tC,='>"?4 m  4nr3rarG)4rZtypingrnumpyr.image_processing_utilsrrrimage_transformsrr image_utilsr r r r rrrrrrrrrrprocessing_utilsrutilsrrrutils.import_utilsrr^modeling_supergluerr>rr get_loggerrWrzr<rr2r:rSrUra__all__rHr3r1rs, UUC -;;*C$   H % 8< a :: a--4 a"8<" "--4""J$J$8L :H0HHV % %r3