Encode image using encoder ONNX model.

Remove overlapping boxes from layout detection results based on a given overlap ratio. Args: layout_det_res (Dict[str, List[Dict]]):

Forward pass for a single stream (i.e. content or query) tgt_norm is just a LayerNorm'd tgt. Added as a separate parameter for effici

Forward pass for a single stream (i.e. content or query) tgt_norm is just a LayerNorm'd tgt. Added as a separate parameter for effici

Args: feat (Tensor): Tensor of shape :math:`(N, D_i, H, W)`. out_enc (Tensor): Encoder output of shape

Generate a square mask for the sequence. The masked positions are filled with float('-inf'). Unmasked positions are filled with float

Generate a square mask for the sequence. The masked positions are filled with float('-inf'). Unmasked positions are filled with float

return metrics { 'precision': 0, 'recall': 0, 'hmean': 0 }

Return set of symbol pairs in a word. Word is represented as tuple of symbols (symbols being variable-length strings).

Get or initialize OpenDocONNX pipeline. Args: layout_model_path: Path to layout detection ONNX model. unirec_encoder_path: Path t

This helper reads input from txt files

This helper determine the fraction of intersection area over detection area

Padding an image using OpenCV: - If the image is smaller than the target size, pad it to 640x640. - If the image is larger than the targe

Convert k points to 1 polygon. Args: points (ndarray or list): A ndarray or a list of shape (2k) that indicates k points.

Calculate the intersection area between two polygon. Args: poly_det (Polygon): A polygon predicted by detector. poly_gt (Polygon)

保存结果为JSON

保存结果为Markdown，按阅读顺序包含图片

Generate shrink_quad_along_width.

Generate shrink_quad_along_width.

f_direction: h x w x 2 pos_list: [[y, x], [y, x], [y, x] ...] binary_tcl_map: h x w

f_direction: h x w x 2 pos_list: [[y, x], [y, x], [y, x] ...]

Sort text boxes in order from top to bottom, left to right args: dt_boxes(array):detected text boxes with shape [4, 2] return:

Test OpenOCR using command-line interface

Test OpenOCR using Python API

Convert the input to a numpy array. Args: img (PIL.Image or numpy.ndarray): Input image. Returns: numpy.ndarray: Numpy

Returns the tokenized representation of given input string(s) Parameters ---------- texts : Union[str, List[str]] An input s

Replace figures in a table area with tokens, return new image and token map. Args: table_block_img (np.ndarray): Table image.

Detect and truncate character-level, phrase-level, or line-level repetition in content. Args: content (str): Input text. lin

Replace tokens in a string with their HTML image equivalents. Args: table_res_str (str): Table string with tokens. figure_to

Initializes the LMDB environment.

Initializes the LMDB environment.