feat: add en/el/th ppocr_v5_rec_mobile paddle/onnx/openvino (#546)

* feat: add en/el/th ppocr_v5_rec_mobile paddle/onnx/openvino

* test: change to a more secure code

Author: SWHL

python/demo.py

@@ -6,7 +6,7 @@
 engine = RapidOCR()
 
 img_url = "https://img1.baidu.com/it/u=3619974146,1266987475&fm=253&fmt=auto&app=138&f=JPEG?w=500&h=516"
-result = engine(img_url, return_word_box=True, return_single_char_box=True)
+result = engine(img_url)
 print(result)
 
 result.vis("vis_result.jpg")

python/rapidocr/default_models.yaml

@@ -29,6 +29,8 @@ class LangRec(Enum):
     TA = "ta"
     TE = "te"
     ESLAV = "eslav"
+    TH = "th"
+    EL = "el"
 
 
 class OCRVersion(Enum):

python/rapidocr/utils/typings.py

@@ -0,0 +1,81 @@
+# -*- encoding: utf-8 -*-
+# @Author: SWHL
+# @Contact: liekkaskono@163.com
+import shlex
+import sys
+from pathlib import Path
+
+import pytest
+from pytest import mark
+
+root_dir = Path(__file__).resolve().parent.parent
+sys.path.append(str(root_dir))
+
+from rapidocr import RapidOCR
+from rapidocr.main import main
+
+tests_dir = root_dir / "tests" / "test_files"
+img_path = tests_dir / "ch_en_num.jpg"
+
+
+@pytest.fixture()
+def engine():
+    engine = RapidOCR()
+    return engine
+
+
+@mark.parametrize("cmd,gt", [(f"--img_path {img_path}", "正品促销")])
+def test_cli(capsys, cmd, gt):
+    main(shlex.split(cmd))
+    output = capsys.readouterr().out.strip()
+    assert gt in output
+
+
+@mark.parametrize("cmd", [f"config --save_cfg_file {tests_dir}/config.yaml"])
+def test_cli_config(capsys, cmd):
+    main(shlex.split(cmd))
+    output = capsys.readouterr().out.strip()
+
+    assert "The config file has saved in" in output
+
+    cfg_yaml_path = tests_dir / "config.yaml"
+    assert cfg_yaml_path.exists()
+    cfg_yaml_path.unlink()
+
+
+@mark.parametrize("cmd", ["check"])
+def test_cli_check(capsys, cmd):
+    main(shlex.split(cmd))
+    output = capsys.readouterr().out.strip()
+
+    assert "Success! rapidocr is installed correctly!" in output
+
+
+@mark.parametrize(
+    "cmd,img_name",
+    [
+        (
+            f"--img_path {img_path} -vis --vis_save_dir {tests_dir}",
+            f"{img_path.stem}_vis.png",
+        ),
+        (
+            f"--img_path {img_path} -vis --vis_save_dir {tests_dir} -word",
+            f"{img_path.stem}_vis_single.png",
+        ),
+    ],
+)
+def test_cli_vis(cmd, img_name):
+    main(shlex.split(cmd))
+    vis_path = tests_dir / img_name
+    assert vis_path.exists()
+    vis_path.unlink()
+
+
+def test_cli_lang_type():
+    img_path = tests_dir / "japan.jpg"
+    cmd = f"--img_path {img_path} --lang_type japan -vis --vis_save_dir {tests_dir}"
+    vis_path = tests_dir / f"{img_path.stem}_vis.png"
+
+    main(shlex.split(cmd))
+    assert vis_path.exists()
+    vis_path.unlink()

python/tests/test_cli.py

@@ -0,0 +1,65 @@
+# -*- encoding: utf-8 -*-
+# @Author: SWHL
+# @Contact: liekkaskono@163.com
+import sys
+from pathlib import Path
+
+import cv2
+import pytest
+
+root_dir = Path(__file__).resolve().parent.parent
+sys.path.append(str(root_dir))
+
+from rapidocr import RapidOCR
+
+tests_dir = root_dir / "tests" / "test_files"
+img_path = tests_dir / "ch_en_num.jpg"
+
+
+@pytest.fixture()
+def engine():
+    engine = RapidOCR()
+    return engine
+
+
+def test_only_det(engine):
+    result = engine(img_path, use_det=True, use_cls=False, use_rec=False)
+    assert len(result) == 18
+
+
+def test_only_cls(engine):
+    img_path = tests_dir / "text_cls.jpg"
+    result = engine(img_path, use_det=False, use_cls=True, use_rec=False)
+    assert len(result) == 1
+    assert result.cls_res[0][0] == "0"
+
+
+def test_only_rec(engine):
+    img_path = tests_dir / "text_rec.jpg"
+    result = engine(img_path, use_det=False, use_cls=False, use_rec=True)
+    assert len(result) == 1
+    assert result.txts[0] == "韩国小馆"
+
+
+def test_det_rec(engine):
+    result = engine(img_path, use_det=True, use_cls=False, use_rec=True)
+    assert len(result) == 18
+    assert result.txts[0] == "正品促销"
+
+
+def test_cls_rec(engine):
+    img_path = tests_dir / "text_cls.jpg"
+    result = engine(img_path, use_det=False, use_cls=True, use_rec=True)
+
+    assert result is not None
+    assert len(result) == 1
+    assert result.txts[0] == "韩国小馆"
+
+
+def test_det_cls_rec(engine):
+    img = cv2.imread(str(img_path))
+
+    result = engine(img)
+    assert result is not None
+    assert len(result) == 18
+    assert result.txts[0] == "正品促销"

python/tests/test_det_cls_rec.py

@@ -0,0 +1,96 @@
+# -*- encoding: utf-8 -*-
+# @Author: SWHL
+# @Contact: liekkaskono@163.com
+import sys
+from pathlib import Path
+
+from pytest import mark
+
+root_dir = Path(__file__).resolve().parent.parent
+sys.path.append(str(root_dir))
+
+from rapidocr import EngineType, ModelType, OCRVersion, RapidOCR
+
+tests_dir = root_dir / "tests" / "test_files"
+img_path = tests_dir / "ch_en_num.jpg"
+
+
+@mark.parametrize(
+    "engine_type",
+    [EngineType.ONNXRUNTIME, EngineType.PADDLE, EngineType.OPENVINO, EngineType.TORCH],
+)
+def test_ppocrv5_rec_mobile(engine_type):
+    engine = RapidOCR(
+        params={
+            "Rec.ocr_version": OCRVersion.PPOCRV5,
+            "Rec.model_type": ModelType.MOBILE,
+            "Rec.engine_type": engine_type,
+        }
+    )
+    img_path = tests_dir / "text_rec.jpg"
+    result = engine(img_path, use_det=False, use_cls=False, use_rec=True)
+
+    assert result.txts is not None
+    assert result.txts[0] == "韩国小馆"
+
+
+@mark.parametrize(
+    "engine_type",
+    [EngineType.ONNXRUNTIME, EngineType.PADDLE, EngineType.OPENVINO, EngineType.TORCH],
+)
+def test_ppocrv5_det_mobile(engine_type):
+    engine = RapidOCR(
+        params={
+            "Det.ocr_version": OCRVersion.PPOCRV5,
+            "Det.model_type": ModelType.MOBILE,
+            "Det.engine_type": engine_type,
+        }
+    )
+    img_path = tests_dir / "ch_en_num.jpg"
+    result = engine(img_path, use_det=True, use_cls=False, use_rec=False)
+
+    assert result.boxes is not None
+    assert len(result.boxes) == 17
+
+
+@mark.skipif(sys.platform.startswith("darwin"), reason="does not run on macOS")
+def test_engine_openvino():
+    engine = RapidOCR(
+        params={
+            "Det.engine_type": EngineType.OPENVINO,
+            "Cls.engine_type": EngineType.OPENVINO,
+            "Rec.engine_type": EngineType.OPENVINO,
+        }
+    )
+
+    result = engine(img_path)
+    assert result.txts is not None
+    assert result.txts[0] == "正品促销"
+
+
+def test_engine_paddle():
+    engine = RapidOCR(
+        params={
+            "Det.engine_type": EngineType.PADDLE,
+            "Cls.engine_type": EngineType.PADDLE,
+            "Rec.engine_type": EngineType.PADDLE,
+        }
+    )
+
+    result = engine(img_path)
+    assert result.txts is not None
+    assert result.txts[0] == "正品促销"
+
+
+def test_engine_torch():
+    engine = RapidOCR(
+        params={
+            "Det.engine_type": EngineType.TORCH,
+            "Cls.engine_type": EngineType.TORCH,
+            "Rec.engine_type": EngineType.TORCH,
+        }
+    )
+
+    result = engine(img_path)
+    assert result.txts is not None
+    assert result.txts[0] == "正品促销"