Sourabhbhalke / Recognizing_Shapes_cv2

Automated detection and recognition of shapes like lines and circles in images using OpenCV, with optimized parameters for diverse image sets.

Recognizing_Shapes_cv2

Automated detection and recognition of shapes like lines and circles in images using OpenCV, with optimized parameters for diverse image sets.

Project Structure

Main Branch

The main branch contains the optimized code for the subway image. This serves as the base implementation, and the code includes:

• Grayscale conversion
• Canny edge detection
• Hough Line Transform
• Hough Circle Transform
• Gaussian Blur
• Optimized parameter settings for shape detection in the subway image

Branches for Other Images

Each of the following images has its own branch, where the code is being optimized for the respective image:

• breakfast (breakfast_branch)
• dinner (dinner_branch)
• building (building_branch)

These branches are forks of the main branch, and contain specific adjustments to the following components:

• Grayscale settings
• Edge detection thresholds
• Hough Transform parameters (for lines and circles)
• Gaussian Blur settings

Workflow

1. Main Branch: Start with the base code optimized for the subway image.
2. Image Branches: For each image branch:
   • Checkout the branch.
   • Adjust parameters specific to the image.
   • Test and refine the code to optimize shape detection.
3. Final Integration: Optionally, merge back the optimized code from the branches into the main branch or a new branch if a generalized solution is developed.

How to Use

1. Checkout the branch for the image you want to work with:

   git checkout breakfast_branch

2. Run the script to detect shapes with optimized parameters for that image.

3. To switch to another image:

   git checkout <image_branch>

Contribution

Feel free to fork the project and submit pull requests with improvements, especially in optimizing detection parameters for new images.