apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-08 10:00:29 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Model] Support PP-StructureV2-Layout model (#1867)

Browse Source 
* [Model] init pp-structurev2-layout code

* [Model] init pp-structurev2-layout code

* [Model] init pp-structurev2-layout code

* [Model] add structurev2_layout_preprocessor

* [PP-StructureV2] add postprocessor and layout detector class

* [PP-StructureV2] add postprocessor and layout detector class

* [PP-StructureV2] add postprocessor and layout detector class

* [PP-StructureV2] add postprocessor and layout detector class

* [PP-StructureV2] add postprocessor and layout detector class

* [pybind] add pp-structurev2-layout model pybind

* [pybind] add pp-structurev2-layout model pybind

* [Bug Fix] fixed code style

* [examples] add pp-structurev2-layout c++ examples

* [PP-StructureV2] add python example and docs

* [benchmark] add pp-structurev2-layout benchmark support
This commit is contained in:
DefTruth
2023-05-05 13:05:58 +08:00
committed by GitHub
parent 2c5fd91a7f
commit 6d0261e9e4
26 changed files with 1255 additions and 23 deletions
Download Patch File Download Diff File Expand all files Collapse all files

112
python/fastdeploy/vision/ocr/ppocr/__init__.py
View File

@@ -650,7 +650,7 @@ class Recognizer(FastDeployModel):
class StructureV2TablePreprocessor:
def __init__(self):
"""Create a preprocessor for StructureV2TableModel
"""Create a preprocessor for StructureV2Table Model
"""
self._preprocessor = C.vision.ocr.StructureV2TablePreprocessor()
@@ -664,12 +664,12 @@ class StructureV2TablePreprocessor:
class StructureV2TablePostprocessor:
def __init__(self):
"""Create a postprocessor for StructureV2TableModel
"""Create a postprocessor for StructureV2Table Model
"""
self._postprocessor = C.vision.ocr.StructureV2TablePostprocessor()
def run(self, runtime_results):
"""Postprocess the runtime results for StructureV2TableModel
"""Postprocess the runtime results for StructureV2Table Model
:param: runtime_results: (list of FDTensor or list of pyArray)The output FDTensor results from runtime
:return: list of Result(If the runtime_results is predict by batched samples, the length of this list equals to the batch size)
"""
@@ -683,10 +683,11 @@ class StructureV2Table(FastDeployModel):
table_char_dict_path="",
runtime_option=None,
model_format=ModelFormat.PADDLE):
"""Load OCR StructureV2Table model provided by PaddleOCR.
"""Load StructureV2Table model provided by PP-StructureV2.
:param model_file: (str)Path of model file, e.g ./ch_ppocr_mobile_v2.0_cls_infer/model.pdmodel.
:param params_file: (str)Path of parameter file, e.g ./ch_ppocr_mobile_v2.0_cls_infer/model.pdiparams, if the model format is ONNX, this parameter will be ignored.
:param table_char_dict_path: (str)Path of table_char_dict file, e.g ../ppocr/utils/dict/table_structure_dict_ch.txt
:param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU.
:param model_format: (fastdeploy.ModelForamt)Model format of the loaded model.
"""
@@ -703,8 +704,8 @@ class StructureV2Table(FastDeployModel):
self._runnable = True
def clone(self):
"""Clone OCR StructureV2Table model object
:return: a new OCR StructureV2Table model object
"""Clone StructureV2Table model object
:return: a new StructureV2Table model object
"""
class StructureV2TableClone(StructureV2Table):
@@ -749,6 +750,105 @@ class StructureV2Table(FastDeployModel):
self._model.postprocessor = value
class StructureV2LayoutPreprocessor:
def __init__(self):
"""Create a preprocessor for StructureV2Layout Model
"""
self._preprocessor = C.vision.ocr.StructureV2LayoutPreprocessor()
def run(self, input_ims):
"""Preprocess input images for StructureV2Layout Model
:param: input_ims: (list of numpy.ndarray)The input image
:return: list of FDTensor
"""
return self._preprocessor.run(input_ims)
class StructureV2LayoutPostprocessor:
def __init__(self):
"""Create a postprocessor for StructureV2Layout Model
"""
self._postprocessor = C.vision.ocr.StructureV2LayoutPostprocessor()
def run(self, runtime_results):
"""Postprocess the runtime results for StructureV2Layout Model
:param: runtime_results: (list of FDTensor or list of pyArray)The output FDTensor results from runtime
:return: list of Result(If the runtime_results is predict by batched samples, the length of this list equals to the batch size)
"""
return self._postprocessor.run(runtime_results)
class StructureV2Layout(FastDeployModel):
def __init__(self,
model_file="",
params_file="",
runtime_option=None,
model_format=ModelFormat.PADDLE):
"""Load StructureV2Layout model provided by PP-StructureV2.
:param model_file: (str)Path of model file, e.g ./picodet_lcnet_x1_0_fgd_layout_infer/model.pdmodel.
:param params_file: (str)Path of parameter file, e.g ./picodet_lcnet_x1_0_fgd_layout_infer/model.pdiparams, if the model format is ONNX, this parameter will be ignored.
:param runtime_option: (fastdeploy.RuntimeOption)RuntimeOption for inference this model, if it's None, will use the default backend on CPU.
:param model_format: (fastdeploy.ModelForamt)Model format of the loaded model.
"""
super(StructureV2Layout, self).__init__(runtime_option)
if (len(model_file) == 0):
self._model = C.vision.ocr.StructureV2Layout()
self._runnable = False
else:
self._model = C.vision.ocr.StructureV2Layout(
model_file, params_file, self._runtime_option, model_format)
assert self.initialized, "StructureV2Layout model initialize failed."
self._runnable = True
def clone(self):
"""Clone StructureV2Layout model object
:return: a new StructureV2Table model object
"""
class StructureV2LayoutClone(StructureV2Layout):
def __init__(self, model):
self._model = model
clone_model = StructureV2LayoutClone(self._model.clone())
return clone_model
def predict(self, input_image):
"""Predict an input image
:param input_image: (numpy.ndarray)The input image data, 3-D array with layout HWC, BGR format
:return: bboxes
"""
if self._runnable:
return self._model.predict(input_image)
return False
def batch_predict(self, images):
"""Predict a batch of input image
:param images: (list of numpy.ndarray) The input image list, each element is a 3-D array with layout HWC, BGR format
:return: list of bboxes list
"""
if self._runnable:
return self._model.batch_predict(images)
return False
@property
def preprocessor(self):
return self._model.preprocessor
@preprocessor.setter
def preprocessor(self, value):
self._model.preprocessor = value
@property
def postprocessor(self):
return self._model.postprocessor
@postprocessor.setter
def postprocessor(self, value):
self._model.postprocessor = value
class PPOCRv3(FastDeployModel):
def __init__(self, det_model=None, cls_model=None, rec_model=None):
"""Consruct a pipeline with text detector, direction classifier and text recognizer models

9
python/fastdeploy/vision/visualize/__init__.py
View File

@@ -23,7 +23,9 @@ def vis_detection(im_data,
labels=[],
score_threshold=0.0,
line_size=1,
font_size=0.5):
font_size=0.5,
font_color=[255, 255, 255],
font_thickness=1):
"""Show the visualized results for detection models
:param im_data: (numpy.ndarray)The input image data, 3-D array with layout HWC, BGR format
@@ -32,10 +34,13 @@ def vis_detection(im_data,
:param score_threshold: (float) score_threshold threshold for result scores, the bounding box will not be shown if the score is less than score_threshold
:param line_size: (float) line_size line size for bounding boxes
:param font_size: (float) font_size font size for text
:param font_color: (list of int) font_color for text
:param font_thickness: (int) font_thickness for text
:return: (numpy.ndarray) image with visualized results
"""
return C.vision.vis_detection(im_data, det_result, labels, score_threshold,
line_size, font_size)
line_size, font_size, font_color,
font_thickness)
def vis_perception(im_data,