0hE dZddlZddlZddlZddlZddlmcmZ ddl m Z ddl m Z  ddlmZ ej"ZeEej*ej,ej.ej0ej2ej4dZe j8j:Zej,ej<ej.ej>ej@ejBejDejFdZ$e dg dd Z%d Z&e d d ddZ'e ddddZ(e ddddZ)e dd ddZ*y#e$reZYwxYw#e$rdZdZY,wxYw)z$Utilities related to image handling.N)backend) keras_export)Image)nearestbilinearbicubichammingboxlanczos)rrrarealanczos3lanczos5gaussian mitchellcubicz&keras.preprocessing.image.smart_resize)v1c<t|dk7rtd|dtj|}|jj K|jj dks|jj dkDrtd|jdtj|}|d|d }}|\}}|jj |jd } nd} tj tj ||zd |z d } tj tj ||zd |z d } tj|| } tj|| } tj tj || z d dz d } tj tj || z d dz d } |jj dk(r3tjd | | d g}tjd | | d g}n0tj| | d g}tj| | d g}tj|||}tjj|||}|jj [|jj dk(r|jddd| f|jj dk(r|jdd| ft|tjr|j!S|S)a Resize images to a target size without aspect ratio distortion. Warning: `tf.keras.preprocessing.image.smart_resize` is not recommended for new code. Prefer `tf.keras.layers.Resizing`, which provides the same functionality as a preprocessing layer and adds `tf.RaggedTensor` support. See the [preprocessing layer guide]( https://www.tensorflow.org/guide/keras/preprocessing_layers) for an overview of preprocessing layers. TensorFlow image datasets typically yield images that have each a different size. However, these images need to be batched before they can be processed by TF-Keras layers. To be batched, images need to share the same height and width. You could simply do: ```python size = (200, 200) ds = ds.map(lambda img: tf.image.resize(img, size)) ``` However, if you do this, you distort the aspect ratio of your images, since in general they do not all have the same aspect ratio as `size`. This is fine in many cases, but not always (e.g. for GANs this can be a problem). Note that passing the argument `preserve_aspect_ratio=True` to `resize` will preserve the aspect ratio, but at the cost of no longer respecting the provided target size. Because `tf.image.resize` doesn't crop images, your output images will still have different sizes. This calls for: ```python size = (200, 200) ds = ds.map(lambda img: smart_resize(img, size)) ``` Your output images will actually be `(200, 200)`, and will not be distorted. Instead, the parts of the image that do not fit within the target size get cropped out. The resizing process is: 1. Take the largest centered crop of the image that has the same aspect ratio as the target size. For instance, if `size=(200, 200)` and the input image has size `(340, 500)`, we take a crop of `(340, 340)` centered along the width. 2. Resize the cropped image to the target size. In the example above, we resize the `(340, 340)` crop to `(200, 200)`. Args: x: Input image or batch of images (as a tensor or NumPy array). Must be in format `(height, width, channels)` or `(batch_size, height, width, channels)`. size: Tuple of `(height, width)` integer. Target size. interpolation: String, interpolation to use for resizing. Supports `bilinear`, `nearest`, `bicubic`, `area`, `lanczos3`, `lanczos5`, `gaussian`, `mitchellcubic`. Defaults to `'bilinear'`. Returns: Array with shape `(size[0], size[1], channels)`. If the input image was a NumPy array, the output is a NumPy array, and if it was a TF tensor, the output is a TF tensor. z6Expected `size` to be a tuple of 2 integers, but got: .NzExpected an image array with shape `(height, width, channels)`, or `(batch_size, height, width, channels)`, but got input with incorrect rank, of shape float32int32r)imagessizemethod)len ValueErrortfconvert_to_tensorshaperankcastminimumstacksliceimageresize set_shape isinstancenpndarraynumpy)xr interpolationimgr#heightwidth target_height target_widthstatic_num_channels crop_height crop_widthcrop_box_hstartcrop_box_wstartcrop_box_start crop_box_sizes X/var/www/html/engine/venv/lib/python3.12/site-packages/tf_keras/src/utils/image_utils.py smart_resizer?BsD 4yA~DTF! L    q !C yy~~! 99>>A !!3;;>99+QH  HHSME"IuRyEF"&M< yy~~!!iim"''  %y1L@'K %y1MA7J **V[1KE:.Jgg $i014gOgg  "I.2GO yy~~1o"JK"k:r!BC ?OQ"GH+z2!>? ((3 6C ((//4 / FC yy~~! 99>>Q  MM4t-@A B 99>>Q  MM4':; <!RZZ yy{ Jc|j}|tvr-tdj|tj t|S)NzFValue not recognized for `interpolation`: {}. Supported values are: {})lower_TF_INTERPOLATION_METHODSNotImplementedErrorformatkeys)r1s r>get_interpolationrGsN!'')M55! f],E,J,J,LM   %] 33r@zkeras.utils.array_to_imgz&keras.preprocessing.image.array_to_imgc|tj}|tj}t t dt j ||}|jdk7rtd|j|dvrtd||dk(r|jd d d }|r<|t j|z }t j|}|d k7r||z}|d z}|jd d k(r%tj|jddS|jd dk(r%tj|jddS|jd d k(rvt j|d kDr/tj|ddddd fjddStj|ddddd fjddStd|jd )a$Converts a 3D Numpy array to a PIL Image instance. Usage: ```python from PIL import Image img = np.random.random(size=(100, 100, 3)) pil_img = tf.keras.utils.array_to_img(img) ``` Args: x: Input data, in any form that can be converted to a Numpy array. data_format: Image data format, can be either `"channels_first"` or `"channels_last"`. None means the global setting `tf.keras.backend.image_data_format()` is used (unless you changed it, it uses `"channels_last"`). Defaults to `None`. scale: Whether to rescale the image such that minimum and maximum values are 0 and 255 respectively. Defaults to `True`. dtype: Dtype to use. None makes the global setting `tf.keras.backend.floatx()` to be used (unless you changed it, it uses `"float32"`). Defaults to `None`. Returns: A PIL Image instance. Raises: ImportError: if PIL is not available. ValueError: if invalid `x` or `data_format` is passed. NzCCould not import PIL.Image. The use of `array_to_img` requires PIL.dtyperzJExpected image array to have rank 3 (single image). Got array with shape: > channels_lastchannels_firstzInvalid data_format: rLrrruint8RGBARGBrILzUnsupported channel number: )rimage_data_formatfloatx pil_image ImportErrorr-asarrayndimr r# transposeminmax fromarrayastype)r0 data_formatscalerJx_maxs r> array_to_imgrbsF//1  }  6   1E"Avv{ %%&WWI /  ==0 >?? && KK1a  q Mq  A: JA SwwqzQ""188G#4f== q""188G#4e<< q 66!9s?&&qAqz'8'8'A3G G""1Q1W:#4#4W#=sCC7 |DEEr@zkeras.utils.img_to_arrayz&keras.preprocessing.image.img_to_arrayc:|tj}|tj}|dvrtd|t j ||}t |jdk(r|dk(r|jddd}|St |jdk(re|dk(r0|jd|jd|jdf}|S|j|jd|jddf}|Std |j) a{Converts a PIL Image instance to a Numpy array. Usage: ```python from PIL import Image img_data = np.random.random(size=(100, 100, 3)) img = tf.keras.utils.array_to_img(img_data) array = tf.keras.utils.image.img_to_array(img) ``` Args: img: Input PIL Image instance. data_format: Image data format, can be either `"channels_first"` or `"channels_last"`. None means the global setting `tf.keras.backend.image_data_format()` is used (unless you changed it, it uses `"channels_last"`). Defaults to `None`. dtype: Dtype to use. None makes the global setting `tf.keras.backend.floatx()` to be used (unless you changed it, it uses `"float32"`). Defaults to `None`. Returns: A 3D Numpy array. Raises: ValueError: if invalid `img` or `data_format` is passed. >rKrLzUnknown data_format: rIrrLrrrMzUnsupported image shape: ) rrTrUr r-rXrr#rZreshape)r2r_rJr0s r> img_to_arrayresB//1  } ==0 >?? 3e$A 177|q * * Aq!$A H QWW  * * 1aggaj!''!*56A H 1771:qwwqz156A H4QWWI>??r@zkeras.utils.save_imgz"keras.preprocessing.image.save_imgc |tj}t|||}|jdk(r0|dk(s|dk(r&t j d|j d}|j|fd|i|y) ajSaves an image stored as a Numpy array to a path or file object. Args: path: Path or file object. x: Numpy array. data_format: Image data format, either `"channels_first"` or `"channels_last"`. file_format: Optional file format override. If omitted, the format to use is determined from the filename extension. If a file object was used instead of a filename, this parameter should always be used. scale: Whether to rescale image values to be within `[0, 255]`. **kwargs: Additional keyword arguments passed to `PIL.Image.save()`. N)r_r`rPjpgjpegz?The JPG format does not support RGBA images, converting to RGB.rQrE)rrTrbmodewarningswarnconvertsave)pathr0r_ file_formatr`kwargsr2s r>save_imgrqRsw//1 qk ?C xx6{e3{f7L M kk%  CHHT0+00r@zkeras.utils.load_imgz"keras.preprocessing.image.load_imgc |rtjdd}t tdt |t j rtj|}nt |tjttfrt |tjrt|j}t|d5}tjt j |j}dddntdt||dk(r j dvri|j#d}nW|d k(r!j d k7rC|j#d }n1|d k(r!j d k7r|j#d }n t%d ||d|df}|j&|k7r|t(vrload_imgrlsX J !  M  $ #nnT" D7<<4 5 dGLL )t||~&D $  7..AFFH!56C 7 7:4:, G  [  88- -++c"C v  88v ++f%C u  88u ++e$CKLL)!nk!n= 88) )$>> %%+V% "<"A"A"CD&2-@H # v.@+ m$}4E $|3 E "&+6   3 #)K#7A"=#(:#5!"; /* < /+ = ##!!  jj!3X8jL Jjj!3X> Ju 7 7s ?7JJ )r)NTN)NN)NNT)FrvNrF)+__doc__rwrzrjr/r-tensorflow.compat.v2compatv2r! tf_keras.srcr tensorflow.python.util.tf_exportrPILrrV Resamplingpil_image_resamplingAttributeErrorrWNEARESTBILINEARBICUBICHAMMINGBOXLANCZOSrr) ResizeMethodNEAREST_NEIGHBORAREALANCZOS3LANCZOS5GAUSSIAN MITCHELLCUBICrCr?rGrbrerqrr@r>rs+ !! :  &)(33#'//(11'//'//#'''// "xx$$ %%,,##   %%%%%%!// 62>y?yx4 HKFKF\ H2 2 j$&JK1L12$&JK sLsU )() I s.E EEEEE E)(E)