ICE2607 Lab2: Canny Edge Detection

This project simplifies Canny edge detection with a command-line interface and customizable parameters. It also incorporates the Otsu method for adaptive thresholding.

File Structure

• codes/
  • images.py: Contains the Image class.
  • main.py: Main Python script.
  • modules.py: Contains processing modules.
  • utils.py: Contains convolution and Otsu functions.
• tests/
  • opencv_figs: Edge detection using OpenCV.
  • step_figs: Show images after each step.
• images/: Contains default test images.
• examples/: Contains edge detection results of default tests.
• report_figures/: Contains additional figures used in the report.
• Lab2_report.pdf
• README.md
• Makefile
• requirements.txt

Example

Install

1. Ensure that you have Python installed on your system.

2. Clone or download this project and navigate to the project directory.

3. Check if the dependencies listed in requirements.txt are installed. If not, create a virtual environment (if necessary) and run make install.

Usage

Default Canny Edge Detection

python -m codes.main \
    --input-dir path/to/input/folder \
    --output-dir path/to/output/folder \
    --output-type png

Note:

• input-dir can be path to the input image or the folder containing input images.
• Avoid duplicate filenames in input-dir.
• If the parameters are not specified, the program will utilize the three images in the ./images directory as input and save all results in PNG format to the ./edges directory.

Personalize Parameters

python -m codes.main \
    --input-dir path/to/input/folder \
    --output-dir path/to/output/folder \
    --operator Sobel \
    --th-low 40 \
    --th-high 100 \
    --sigma 0.2

Note:

• The operator can be selected from 'Roberts', 'Sobel', 'Prewitt', and 'Canny'.
• Avoid duplicate filenames in input-dir.
• th-low will be set to 0.4 * th-high if not specified.

Adaptive Thresholding

python -m codes.main \
    --input-dir path/to/input/folder \
    --output-dir path/to/output/folder \
    --otsu True

Acknowledgements

https://github.com/khushitejwani/Canny-Edge-Detection-Using-Otsu-Thresholding