chore: init

.gitattributes

@@ -0,0 +1,2 @@
+# Auto detect text files and perform LF normalization
+* text=auto

.gitignore

@@ -0,0 +1,156 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos 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/
+cover/
+
+# 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
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .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
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.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
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintainted in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+.DS_Store
+
+data

.vscode/settings.json

@@ -0,0 +1,6 @@
+{
+    "python.formatting.provider": "black",
+    "[python]": {
+        "editor.formatOnSave": true
+    }
+}

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2023 JacobLinCool
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -0,0 +1 @@
+# captcha-recognizer

pyproject.toml

@@ -0,0 +1,27 @@
+[tool.poetry]
+name = "captcha-recognizer"
+version = "0.0.0"
+description = ""
+authors = ["JacobLinCool <[email protected]>"]
+license = "MIT"
+readme = "README.md"
+
+[tool.poetry.dependencies]
+python = "~3.10"
+opencv-python = "^4.7.0.68"
+gradio = "^3.18.0"
+setuptools = "^67.3.2"
+pytesseract = "^0.3.10"
+
+[tool.poetry.group.dev.dependencies]
+black = "^23.1.0"
+poethepoet = "^0.18.1"
+
+[tool.poe.tasks]
+format = "black ."
+collect = "python -m scripts/collect"
+preprocess = "python -m scripts.preprocess"
+
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"

scripts/collect.py

@@ -0,0 +1,24 @@
+# Description: Collects sample images for the website
+import os
+import time
+import urllib.request
+from src.shared import raw_dir
+
+count = 100
+
+
+def main():
+    print(f"Collecting {count} images to {raw_dir}")
+
+    for i in range(count):
+        url = "https://cos2s.ntnu.edu.tw/AasEnrollStudent/RandImage"
+        filename = os.path.join(raw_dir, f"{i}.jpg")
+        urllib.request.urlretrieve(url, filename)
+        print(f"Downloaded {i+1}/{count} {filename}")
+        time.sleep(0.1)
+
+    print("Done")
+
+
+if __name__ == "__main__":
+    main()

scripts/generator.py

@@ -0,0 +1,93 @@
+import os
+from PIL import Image, ImageFont, ImageDraw
+from random import choice, random
+from string import ascii_lowercase
+from src.shared import genereated_dir
+
+WIDTH, HEIGHT = 108, 30
+
+count = 100
+
+font = ImageFont.truetype(
+    os.path.join(os.path.dirname(__file__), "..", "fonts", "NotoSans-Regular.ttf"), 18
+)
+
+
+def gen():
+    image = Image.new("RGB", (WIDTH, HEIGHT))
+
+    if random() > 0.5:
+        # alphabetical
+        a = choice(ascii_lowercase)
+        b = choice(ascii_lowercase)
+        c = choice(ascii_lowercase)
+        d = choice(ascii_lowercase)
+        ans = a + b + c + d
+
+        image = draw(image, [a, b, c, d])
+
+    else:
+        # arithmetic
+        a = choice(range(10))
+        b = choice(range(10))
+        op = choice(["+", "-", "x"])
+
+        if op == "+":
+            ans = f"{a}+{b}="
+        elif op == "-":
+            ans = f"{a}-{b}="
+        else:
+            ans = f"{a}x{b}="
+
+        if op == "x" and random() > 0.5:
+            op = "X"
+
+        image = draw(image, [str(a), op, str(b), "="])
+
+    return image, str(ans)
+
+
+def draw(image: Image, text: list[str]) -> Image:
+    draw = ImageDraw.Draw(image)
+    draw.rectangle((0, 0, WIDTH, HEIGHT), fill=(255, 255, 255))
+
+    for i, t in enumerate(text):
+        txt = Image.new("RGBA", (30, 30))
+        d = ImageDraw.Draw(txt)
+        d.text(
+            (choice(range(0, 15)), -5 + choice(range(0, 15))),
+            t,
+            font=font,
+            fill=(255, 0, 0),
+        )
+        image.paste(txt, (14 + (i * 20), 0), txt)
+
+    # draw noise lines
+    for i in range(30):
+        fill = choice([120, 200])
+        x = random() * WIDTH
+        y = random() * HEIGHT
+
+        draw.line(
+            (
+                x,
+                y,
+                x + 15 * (random() - 1),
+                y + 15 * (random() - 1),
+            ),
+            fill=(fill, fill, fill),
+            width=1,
+        )
+
+    return image
+
+
+if __name__ == "__main__":
+    for i in range(count):
+        image, ans = gen()
+        image.save(os.path.join(genereated_dir, f"{i}.png"))
+        with open(os.path.join(genereated_dir, f"{i}.txt"), "w") as f:
+            f.write(ans)
+            print(f"Generated {i}.png and {i}.txt")
+
+    print("Done")

scripts/preprocess.py

@@ -0,0 +1,32 @@
+# Description: Preprocesses sample images
+import os
+import cv2
+import numpy as np
+from PIL import Image
+from src.shared import raw_dir, preprocess_dir
+from src.preprocess import preprocess
+
+
+def main():
+    print(f"Preprocessing images in {raw_dir}")
+
+    for filename in os.listdir(raw_dir):
+        if not filename.endswith(".jpg"):
+            continue
+
+        raw_path = os.path.join(raw_dir, filename)
+        image = np.array(Image.open(raw_path))
+
+        image = preprocess(image)
+
+        # Save to preprocessed
+        preprocessed_path = os.path.join(preprocess_dir, filename)
+        cv2.imwrite(preprocessed_path, image)
+
+        print(f"Preprocessed {filename}")
+
+    print("Done")
+
+
+if __name__ == "__main__":
+    main()

src/app.py

@@ -0,0 +1,27 @@
+import gradio as gr
+import numpy as np
+from preprocess import preprocess
+from solve import solve
+
+
+def run(img: np.ndarray) -> tuple[np.ndarray, str]:
+    preprocessed = preprocess(img)
+
+    solved = solve(preprocessed)
+
+    return preprocessed, solved
+
+
+app = gr.Interface(
+    fn=run,
+    inputs=[
+        gr.Image(label="captcha image", shape=(108, 30)),
+    ],
+    outputs=[
+        gr.Image(label="preprocessed", shape=(108, 30)),
+        gr.Textbox(label="solved"),
+    ],
+    allow_flagging="never",
+)
+
+app.queue().launch()

src/preprocess.py

@@ -0,0 +1,28 @@
+import cv2
+import numpy as np
+
+
+def preprocess(image: np.ndarray) -> np.ndarray:
+    # Upscale, interpolation with nearest neighbor
+    image = cv2.resize(image, (0, 0), fx=3, fy=3, interpolation=cv2.INTER_NEAREST)
+
+    # Denoise gray-like pixels
+    hsv = cv2.cvtColor(image, cv2.COLOR_BGR2HSV)
+    mask = cv2.inRange(hsv, (0, 70, 70), (255, 255, 255))
+    mask = cv2.bitwise_not(mask)
+    image[np.where(mask)] = 255
+
+    # Convert to binary
+    image = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
+    _, image = cv2.threshold(image, 0, 255, cv2.THRESH_OTSU)
+
+    # Fix some holes
+    kernel = np.ones((3, 3), np.uint8)
+    image = cv2.morphologyEx(image, cv2.MORPH_OPEN, kernel, iterations=2)
+
+    # add padding
+    image = cv2.copyMakeBorder(
+        image, 5, 5, 5, 5, cv2.BORDER_CONSTANT, value=(255, 255, 255)
+    )
+
+    return image

src/shared.py

@@ -0,0 +1,20 @@
+import os
+
+raw_dir = os.path.normpath(os.path.join(os.path.dirname(__file__), "..", "data", "raw"))
+
+if not os.path.exists(raw_dir):
+    os.makedirs(raw_dir)
+
+preprocess_dir = os.path.normpath(
+    os.path.join(os.path.dirname(__file__), "..", "data", "preprocessed")
+)
+
+if not os.path.exists(preprocess_dir):
+    os.makedirs(preprocess_dir)
+
+genereated_dir = os.path.normpath(
+    os.path.join(os.path.dirname(__file__), "..", "data", "generated")
+)
+
+if not os.path.exists(genereated_dir):
+    os.makedirs(genereated_dir)

src/solve.py

@@ -0,0 +1,52 @@
+import pytesseract
+import numpy as np
+
+
+def solve(image: np.ndarray) -> str:
+    for mode in [7, 10, 11, 12, 13]:
+        result = normalize(
+            pytesseract.image_to_string(
+                image, lang="eng", config=f"--oem 3 --psm {mode}", timeout=0.5
+            ).strip()
+        )
+        if result != "":
+            return result
+
+    return "not sure"
+
+
+def normalize(s: str) -> str:
+    print(s)
+    if "\n" in s:
+        return ""
+
+    s = s.replace(" ", "").lower()
+
+    # if first is number
+    if s[0].isdigit() and s[2].isdigit():
+        if s[1] in ["+", "4"]:
+            return str(int(s[0]) + int(s[2]))
+        elif s[1] in ["-", "_"]:
+            return str(int(s[0]) - int(s[2]))
+        else:
+            return str(int(s[0]) * int(s[2]))
+
+    # possible alphabet mapping
+    mapping = {
+        ")": "l",
+        "¥": "y",
+        "2": "z",
+        "é": "e",
+    }
+
+    for k, v in mapping.items():
+        s = s.replace(k, v)
+
+    # if not all alphabet
+    if not all([c.isalpha() for c in s]):
+        return ""
+
+    if len(s) != 4:
+        return ""
+
+    return s