Image to 3D Mesh Converter

This script converts input images into 3D mesh data represented in JSON format. It generates a mesh by dividing the image into chunks, creating a point cloud for each chunk, and then triangulating the point cloud to form the mesh.

The intended use for this script is to take WoW minimaps [currently broken] or WoW world map images [working], and convert them into 3D meshes. The ideal settings for my tests are as follows: --z_scale 0.12 --flip_y --flip_z

The #5 example shows you how to format the commandline such that you will get a single JSON file with every bit of useful data out of yor input, and provide you a decent low-quality example that shouldn't take too long to process. The #4 example will take longer to process, but will be of double resolution (original input resolution).

Requirements

• Python 3.x
• OpenCV (cv2)
• NumPy
• SciPy
• tqdm

Usage

1. Installation: Make sure you have Python installed. Install the required packages using pip:

   pip install opencv-python numpy scipy tqdm
   

2. Running the Script: Run the script 3dM111.py with the following command-line arguments:

   python 3dM111.py input_path output_dir [--options]
   

   • input_path: Path to the input image or directory containing images.
   • output_dir: Directory to save the generated mesh data and optionally the processed images.
   • Options:
     • --chunk_size: Size of the chunks (default: 256).
     • --overlap: Overlap size between chunks (default: 32).
     • --z_scale: Scale for depth values (default: 1.0).
     • --downsample: Downsample input images (default: False).
     • --positive_z_only: Disable negative Z values (default: False).
     • --flip_x: Flip X coordinates (default: False).
     • --flip_y: Flip Y coordinates (default: False).
     • --flip_z: Flip Z coordinates (default: False).
     • --export_processed_image: Export processed image (default: False).

3. Output: The script generates JSON files containing mesh data for each input image in the specified output directory. Optionally, processed images (blurred and/or downsampled) can be exported alongside the JSON files.

Examples

1. Convert a single image:

   python 3dM111.py input_image.png output_directory --export_processed_image
   

2. Convert all images in a directory:

   python 3dM111.py input_directory output_directory --flip_x --flip_z
   

3. Convert images with specific options:

   python 3dM111.py input_directory output_directory --downsample --positive_z_only
   

4. Convert WoW minimaps and world maps with these specific options [Best Quality]:

   python 3dM111.py input_image.png output_directory --z_scale 0.12 --flip_y --flip_z
   

5. Convert WoW minimaps and world maps with these specific options [Worst Quality]:

   python 3dM111.py input_image.png output_directory --z_scale 0.12 --flip_y --flip_z --downsample
   

Notes

• Ensure input images are in a format supported by OpenCV (e.g., PNG, JPEG).
• Adjust options such as chunk size, overlap, and coordinate flipping based on the desired output.
• Outputs will either be in chunks that the script will automatically create, or will take pre-chunked minimap tiles.

---

Mesh Data Converter

This script converts JSON mesh data into various 3D mesh formats using the Trimesh library.

Usage

python meshConverter11.py input_json output_dir

• input_json: Path to the input JSON file containing mesh data.
• output_dir: Directory to save the exported meshes.

Dependencies

• Python 3.x
• Trimesh library

Installation

You can install Trimesh via pip:

pip install trimesh

Example

python meshConverter11.py mesh_data.json exported_meshes

This will convert the mesh data from mesh_data.json into OBJ, STL (ASCII), and PLY formats, saving them in the exported_meshes directory.