Dimensor is a graphical user interface (GUI) application for estimating depth from images using the MiDaS model and visualizing the results as a 3D point cloud. The application is built with Tkinter for the GUI, Open3D for 3D visualization, and PyTorch for depth estimation.

• Load and display images using a file selection dialog.
• Process images to generate depth maps.
• Visualize depth maps as 3D point clouds.

• Python 3.7 or higher
• PyTorch
• torchvision
• Open3D
• OpenCV
• Pillow

1. Clone the Repository

   git clone https://github.com/yourusername/dimensor.git
   cd dimensor

2. Create and Activate a Virtual Environment

   python -m venv venv
   source venv/bin/activate  # On Windows use `venv\Scripts\activate`

3. Install Required Packages

4. Download MiDaS Model The MiDaS model will be downloaded automatically when you run the application. Ensure you have an internet connection.

1. Run the Application

   python app.py

2. Use the GUI

   • Browse: Click the "Browse" button to open a file dialog and select an image file. The file path will be automatically filled in the entry field.
   • Load Image: Click the "Load Image" button to display the image in the GUI.
   • Process Image: Click the "Process Image" button to generate and visualize the 3D point cloud.

• app.py: Main application file with Tkinter GUI and depth estimation logic.
• depth_estimation.py: Module containing image processing and depth estimation functions.
• requirements.txt: List of required Python packages.

Feel free to submit issues or pull requests if you have suggestions or improvements.

This project is licensed under the MIT License - see the LICENSE file for details.