Update rapidocr_openvino

python/rapidocr_openvino/ch_ppocr_v3_det/text_detect.py

@@ -25,8 +25,31 @@
 
 class TextDetector:
     def __init__(self, config):
-        self.preprocess_op = create_operators(config["pre_process"])
-        self.postprocess_op = DBPostProcess(**config["post_process"])
+        pre_process_list = {
+            "DetResizeForTest": {
+                "limit_side_len": config.get("limit_side_len", 736),
+                "limit_type": config.get("limit_type", "min"),
+            },
+            "NormalizeImage": {
+                "std": [0.229, 0.224, 0.225],
+                "mean": [0.485, 0.456, 0.406],
+                "scale": "1./255.",
+                "order": "hwc",
+            },
+            "ToCHWImage": None,
+            "KeepKeys": {"keep_keys": ["image", "shape"]},
+        }
+        self.preprocess_op = create_operators(pre_process_list)
+
+        post_process = {
+            "thresh": config.get("thresh", 0.3),
+            "box_thresh": config.get("box_thresh", 0.5),
+            "max_candidates": config.get("max_candidates", 1000),
+            "unclip_ratio": config.get("unclip_ratio", 1.6),
+            "use_dilation": config.get("use_dilation", True),
+            "score_mode": config.get("score_mode", "fast"),
+        }
+        self.postprocess_op = DBPostProcess(**post_process)
 
         self.infer = OpenVINOInferSession(config)
 

python/rapidocr_openvino/config.yaml

@@ -7,67 +7,34 @@ Global:
     width_height_ratio: 8
 
 Det:
-    module_name: ch_ppocr_v3_det
-    class_name: TextDetector
-    model_path: models/ch_PP-OCRv3_det_infer.onnx
-
     use_cuda: false
-    # Details of the params: https://onnxruntime.ai/docs/execution-providers/CUDA-ExecutionProvider.html
-    CUDAExecutionProvider:
-        device_id: 0
-        arena_extend_strategy: kNextPowerOfTwo
-        cudnn_conv_algo_search: EXHAUSTIVE
-        do_copy_in_default_stream: true
 
-    pre_process:
-        DetResizeForTest:
-            limit_side_len: 736
-            limit_type: min
-        NormalizeImage:
-            std: [0.229, 0.224, 0.225]
-            mean: [0.485, 0.456, 0.406]
-            scale: 1./255.
-            order: hwc
-        ToCHWImage:
-        KeepKeys:
-            keep_keys: ['image', 'shape']
+    model_path: models/ch_PP-OCRv3_det_infer.onnx
+
+    limit_side_len: 736
+    limit_type: min
 
-    post_process:
-        thresh: 0.3
-        box_thresh: 0.5
-        max_candidates: 1000
-        unclip_ratio: 1.6
-        use_dilation: true
-        score_mode: fast
+    thresh: 0.3
+    box_thresh: 0.5
+    max_candidates: 1000
+    unclip_ratio: 1.6
+    use_dilation: true
+    score_mode: fast
 
 Cls:
-    module_name: ch_ppocr_v2_cls
-    class_name: TextClassifier
-    model_path: models/ch_ppocr_mobile_v2.0_cls_infer.onnx
-
     use_cuda: false
-    CUDAExecutionProvider:
-        device_id: 0
-        arena_extend_strategy: kNextPowerOfTwo
-        cudnn_conv_algo_search: EXHAUSTIVE
-        do_copy_in_default_stream: true
+
+    model_path: models/ch_ppocr_mobile_v2.0_cls_infer.onnx
 
     cls_image_shape: [3, 48, 192]
     cls_batch_num: 6
     cls_thresh: 0.9
     label_list: ['0', '180']
 
 Rec:
-    module_name: ch_ppocr_v3_rec
-    class_name: TextRecognizer
-    model_path: models/ch_PP-OCRv3_rec_infer.onnx
-
     use_cuda: false
-    CUDAExecutionProvider:
-        device_id: 0
-        arena_extend_strategy: kNextPowerOfTwo
-        cudnn_conv_algo_search: EXHAUSTIVE
-        do_copy_in_default_stream: true
+
+    model_path: models/ch_PP-OCRv3_rec_infer.onnx
 
     rec_img_shape: [3, 48, 320]
     rec_batch_num: 6

python/rapidocr_openvino/rapid_ocr_api.py

@@ -4,7 +4,7 @@
 import copy
 import importlib
 from pathlib import Path
-from typing import Union
+from typing import Optional, Union
 
 import cv2
 import numpy as np
@@ -18,10 +18,13 @@
 
 
 class RapidOCR:
-    def __init__(self, **kwargs):
-        config_path = str(root_dir / "config.yaml")
+    def __init__(self, config_path: Optional[str] = None, **kwargs):
+        if config_path is None:
+            config_path = str(root_dir / "config.yaml")
+
         if not Path(config_path).exists():
             raise FileExistsError(f"{config_path} does not exist!")
+
         config = read_yaml(config_path)
         config = concat_model_path(config)
 

python/rapidocr_openvino/utils.py

@@ -4,7 +4,7 @@
 import argparse
 from io import BytesIO
 from pathlib import Path
-from typing import Union
+from typing import Dict, List, Union
 
 import cv2
 import numpy as np
@@ -194,37 +194,52 @@ def update_global_params(self, config, global_dict):
         return config
 
     def update_det_params(self, config, det_dict):
-        if det_dict:
-            det_dict = {k.split("det_")[1]: v for k, v in det_dict.items()}
-            if not det_dict["model_path"]:
-                det_dict["model_path"] = str(root_dir / config["model_path"])
-            config.update(det_dict)
+        if not det_dict:
+            return config
+
+        det_dict = {k.split("det_")[1]: v for k, v in det_dict.items()}
+        model_path = det_dict.get('model_path', None)
+        if not model_path:
+            det_dict["model_path"] = str(root_dir / config["model_path"])
+
+        config.update(det_dict)
         return config
 
     def update_cls_params(self, config, cls_dict):
-        if cls_dict:
-            need_remove_prefix = ["cls_label_list", "cls_model_path"]
-            new_cls_dict = {}
-            for k, v in cls_dict.items():
-                if k in need_remove_prefix:
-                    k = k.split("cls_")[1]
-                new_cls_dict[k] = v
-
-            if not new_cls_dict["model_path"]:
-                new_cls_dict["model_path"] = str(root_dir / config["model_path"])
-            config.update(new_cls_dict)
+        if not cls_dict:
+            return config
+
+        need_remove_prefix = ["cls_label_list", "cls_model_path", "cls_use_cuda"]
+        new_cls_dict = self.remove_prefix(cls_dict, 'cls_', need_remove_prefix)
+
+        model_path = new_cls_dict.get('model_path', None)
+        if model_path:
+            new_cls_dict["model_path"] = str(root_dir / config["model_path"])
+
+        config.update(new_cls_dict)
         return config
 
     def update_rec_params(self, config, rec_dict):
-        if rec_dict:
-            need_remove_prefix = ["rec_model_path"]
-            new_rec_dict = {}
-            for k, v in rec_dict.items():
-                if k in need_remove_prefix:
-                    k = k.split("rec_")[1]
-                new_rec_dict[k] = v
-
-            if not new_rec_dict["model_path"]:
-                new_rec_dict["model_path"] = str(root_dir / config["model_path"])
-            config.update(new_rec_dict)
+        if not rec_dict:
+            return config
+
+        need_remove_prefix = ["rec_model_path", "rec_use_cuda"]
+        new_rec_dict = self.remove_prefix(rec_dict, 'rec_', need_remove_prefix)
+
+        model_path = new_rec_dict.get('model_path', None)
+        if not model_path:
+            new_rec_dict["model_path"] = str(root_dir / config["model_path"])
+
+        config.update(new_rec_dict)
         return config
+
+    @staticmethod
+    def remove_prefix(
+        config: Dict[str, str], prefix: str, remove_params: List[str]
+    ) -> Dict[str, str]:
+        new_rec_dict = {}
+        for k, v in config.items():
+            if k in remove_params:
+                k = k.split(prefix)[1]
+            new_rec_dict[k] = v
+        return new_rec_dict