first commit

.codex/.gitignore

@@ -0,0 +1 @@
+tmp/

.gitignore

@@ -0,0 +1,184 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*.pyc
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+tmp/
+develop-eggs/
+dist/
+wheels_dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+pip-wheel-metadata/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+requirements.txt
+setup.py
+x_anylabeling_cvhub520.egg-info
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other info into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+mlruns/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# Profiling
+*.pclprof
+
+# pyenv
+.python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+.idea
+.github
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+venv-*/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# VSCode project settings
+.vscode/
+.devcontainer/
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# datasets and projects
+/datasets
+data
+runs/
+wandb/
+zipped_models/
+.DS_Store
+
+# Generated UI files
+*_ui.py
+anylabeling/app_info.py-*
+
+# Neural Network weights -----------------------------------------------------------------------------------------------
+weights/
+*.weights
+*.pt
+*.ts
+*.pb
+*.onnx
+*.engine
+*.mlmodel
+*.mlpackage
+*.torchscript
+*.tflite
+*.h5
+*.mnn
+*_saved_model/
+*_web_model/
+*_openvino_model/
+*_paddle_model/
+*_ncnn_model/
+*_imx_model/
+pnnx*
+*.rknn
+
+# calibration image
+calibration_*.npy

classify_images.py

@@ -0,0 +1,197 @@
+"""
+gavin 2026-01-06
+"""
+
+from __future__ import annotations
+
+import argparse
+import shutil
+from datetime import datetime
+from pathlib import Path
+from typing import Iterable, List, Tuple
+
+from PIL import Image, ImageChops, UnidentifiedImageError
+
+
+def is_color_image(image_path: Path, tolerance: int = 3) -> bool:
+    """
+    Return True if the image contains color information beyond grayscale.
+
+    The tolerance parameter allows minor compression artifacts without
+    misclassifying grayscale images as color.
+    """
+    with Image.open(image_path) as img:
+        rgb = img.convert("RGB")
+        r, g, b = rgb.split()
+        diff_rg = ImageChops.difference(r, g)
+        diff_rb = ImageChops.difference(r, b)
+        max_diff = max(diff_rg.getextrema()[1], diff_rb.getextrema()[1])
+        return max_diff > tolerance
+
+
+def iter_image_files(directory: Path) -> List[Path]:
+    return [path for path in sorted(directory.iterdir()) if path.is_file()]
+
+
+def unique_destination(target_dir: Path, filename: str) -> Path:
+    candidate = target_dir / filename
+    if not candidate.exists():
+        return candidate
+
+    stem = candidate.stem
+    suffix = candidate.suffix
+    index = 1
+    while True:
+        candidate = target_dir / f"{stem}_{index}{suffix}"
+        if not candidate.exists():
+            return candidate
+        index += 1
+
+
+def classify_images(
+    input_dir: Path,
+    color_dir: Path,
+    bw_dir: Path,
+    *,
+    move_files: bool = False,
+    tolerance: int = 3,
+    dry_run: bool = False,
+) -> Tuple[int, int, int]:
+    """
+    Classify images in input_dir into color or black/white destinations.
+
+    Returns a tuple of (color_count, bw_count, skipped_count).
+    """
+    color_dir.mkdir(parents=True, exist_ok=True)
+    bw_dir.mkdir(parents=True, exist_ok=True)
+
+    color_count = bw_count = skipped_count = 0
+    image_paths = iter_image_files(input_dir)
+    total = len(image_paths)
+
+    def print_progress(done: int) -> None:
+        if total == 0:
+            return
+        if done % 10 != 0 and done != total:
+            return
+        percent = done / total * 100
+        print(f"处理进度: {done}/{total} ({percent:5.1f}%)")
+
+    processed = 0
+    for image_path in image_paths:
+        try:
+            has_color = is_color_image(image_path, tolerance=tolerance)
+        except (UnidentifiedImageError, OSError) as exc:
+            print(f"跳过无法读取的文件: {image_path} ({exc})")
+            skipped_count += 1
+            continue
+
+        target_dir = color_dir if has_color else bw_dir
+        destination = unique_destination(target_dir, image_path.name)
+
+        if dry_run:
+            action = "彩色" if has_color else "黑白"
+            print(f"[dry-run] {image_path} -> {action} ({destination.name})")
+        else:
+            if move_files:
+                shutil.move(str(image_path), destination)
+            else:
+                shutil.copy2(image_path, destination)
+
+        if has_color:
+            color_count += 1
+        else:
+            bw_count += 1
+
+        processed += 1
+        print_progress(processed)
+
+    return color_count, bw_count, skipped_count
+
+
+def build_parser() -> argparse.ArgumentParser:
+    timestamp = datetime.now().strftime("%Y%m%d-%H%M%S")
+    default_color_dir = Path("data") / f"c-{timestamp}"
+    default_bw_dir = Path("data") / f"b-{timestamp}"
+
+    parser = argparse.ArgumentParser(
+        description="将图片按彩色或黑白分类到指定目录。"
+    )
+    parser.add_argument(
+        "-i",
+        "--input-dir",
+        type=Path,
+        default=Path("data/color/mix"),
+        help="待分类的图片目录（默认: data/color/mix）。",
+    )
+    parser.add_argument(
+        "-c",
+        "--color-dir",
+        type=Path,
+        default=default_color_dir,
+        help=f"彩色图片输出目录（默认: {default_color_dir}）。",
+    )
+    parser.add_argument(
+        "-b",
+        "--bw-dir",
+        type=Path,
+        default=default_bw_dir,
+        help=f"黑白/灰阶图片输出目录（默认: {default_bw_dir}）。",
+    )
+    parser.add_argument(
+        "-t",
+        "--tolerance",
+        type=int,
+        default=3,
+        help="颜色通道差异阈值，越大越宽松（默认: 3）。",
+    )
+    parser.add_argument(
+        "--move",
+        dest="move_files",
+        action="store_true",
+        help="移动文件而非复制，谨慎使用。",
+    )
+    parser.add_argument(
+        "--dry-run",
+        action="store_true",
+        help="仅输出分类结果，不执行复制/移动。",
+    )
+    return parser
+
+
+def exit_with_usage(parser: argparse.ArgumentParser, message: str) -> None:
+    # 额外打印 usage，便于用户看到用法提示
+    parser.print_usage()
+    parser.exit(2, f"{parser.prog}: error: {message}\n")
+
+
+def main() -> None:
+    parser = build_parser()
+    args = parser.parse_args()
+
+    if not args.input_dir.exists():
+        exit_with_usage(parser, f"输入目录不存在: {args.input_dir}")
+    if not args.input_dir.is_dir():
+        exit_with_usage(parser, f"输入路径不是目录: {args.input_dir}")
+    if args.tolerance < 0:
+        exit_with_usage(parser, "tolerance 需为非负整数。")
+
+    color_count, bw_count, skipped_count = classify_images(
+        args.input_dir,
+        args.color_dir,
+        args.bw_dir,
+        move_files=args.move_files,
+        tolerance=args.tolerance,
+        dry_run=args.dry_run,
+    )
+
+    summary = (
+        f"彩色: {color_count} 张，"
+        f"黑白/灰阶: {bw_count} 张，"
+        f"跳过: {skipped_count} 张。"
+    )
+    print(summary)
+
+
+if __name__ == "__main__":
+    main()