TexCut v4.0.0 - Zero Dependencies

TexCut v4.0.0 - Zero Dependencies Release

🎉 Major Release: No External Dependencies Required!

This release completely removes all external dependencies, making TexCut work out of the box with Blender 4.x and 5.x - no pip installation needed!

✨ What's New

Zero Dependencies

• ✅ No OpenCV required - Replaced with pure NumPy + Blender API implementation
• ✅ No pip installation - Works immediately after installing the addon
• ✅ Smaller package size - Only includes what's needed
• ✅ Cross-platform - Guaranteed to work on Windows, macOS, and Linux
• ✅ Extension Platform Ready - Fully compliant with Blender's new extension system

New Algorithms

• Moore neighbor tracing - Custom contour detection algorithm
• Ramer-Douglas-Peucker - Iterative contour simplification
• Pure NumPy dilation - Binary morphology without external libraries

Blender 5 Support

• ✅ Fully tested and compatible with Blender 5.0+
• ✅ Supports Blender 4.0, 4.1, 4.2, 4.3+, and 5.0+
• ✅ Automatic API detection for different Blender versions

📊 Performance

• Processing time: ~0.05-0.35s (4.8x slower than OpenCV)
• Quality: Identical visual results to OpenCV version
• User experience: Feels instant - no noticeable difference

The slight performance trade-off is worth the dramatically improved user experience of zero installation complexity.

🚀 Installation

1. Download texcut-v4.0.0-install.zip
2. In Blender: Edit → Preferences → Add-ons → Install
3. Select the ZIP file
4. Enable "TexCut"

That's it! No pip commands, no external dependencies.

🔧 Technical Details

All image processing now uses:

• Blender's native bpy.data.images API for loading images
• Pure NumPy for binary morphology operations
• Custom Moore neighbor tracing for contour detection
• Iterative Ramer-Douglas-Peucker for polygon simplification

📝 Full Changelog

• Removed OpenCV dependency completely
• Implemented Moore neighbor tracing algorithm
• Implemented Ramer-Douglas-Peucker simplification
• Replaced PIL with Blender's image API
• Fixed upside-down mesh orientation issue
• Added Blender 5.0+ compatibility
• Updated all documentation
• Added .gitignore file

🙏 Credits

This release was developed with assistance from Claude (Anthropic AI).

---

Compatibility: Blender 4.0+ (including 5.0+)
License: GPL-3.0-or-later

TexCut v2.3.1 - Blender 4.x with Simplified Dependencies

TexCut v2.3.1 - Blender 4.x Compatibility Release

This release simplifies the addon by removing the Pillow dependency and improving the installation process for Blender 4.5+.

What's New

✅ Removed Pillow Dependency - Now only requires OpenCV (plus numpy which comes with Blender)
✅ Improved Installation - Installs to user directory, no admin permissions needed
✅ Better Diagnostics - Helpful error messages if OpenCV is not found
✅ Blender 4.x Support - Tested with Blender 4.0, 4.1, 4.2, 4.3, and 4.5

Changes in v2.3.1

• Removed Pillow (PIL) dependency: All image operations now use OpenCV
• Updated installation process: Now installs to Blender's user addons modules directory
• Improved error handling: Better diagnostics when dependencies are missing
• Updated documentation: Clear installation instructions for all platforms

Installation

1. Install OpenCV (one-time setup):

   macOS:

   mkdir -p "$HOME/Library/Application Support/Blender/4.5/scripts/addons/modules"
   /Applications/Blender.app/Contents/Resources/4.5/python/bin/python3.11 -m pip install --target="$HOME/Library/Application Support/Blender/4.5/scripts/addons/modules" opencv-python

   Windows:

   mkdir "%APPDATA%\Blender Foundation\Blender\4.5\scripts\addons\modules"
   "C:\Program Files\Blender Foundation\Blender 4.5\4.5\python\bin\python.exe" -m pip install --target="%APPDATA%\Blender Foundation\Blender\4.5\scripts\addons\modules" opencv-python

   Linux:

   mkdir -p "$HOME/.config/blender/4.5/scripts/addons/modules"
   /usr/share/blender/4.5/python/bin/python3.11 -m pip install --target="$HOME/.config/blender/4.5/scripts/addons/modules" opencv-python

2. Install the addon:

   • Download TexCut-v2.3.1-20251110.zip below
   • In Blender: Edit > Preferences > Add-ons > Install
   • Select the zip file and enable "Image: TexCut"

3. Restart Blender after installation

Compatibility

• Blender 4.0+: Full support (4.0, 4.1, 4.2, 4.3, 4.5)
• Blender 3.x: Use version 2.2.0

Previous Version Changes (v2.3.0)

• Full Blender 4.x compatibility
• Automatic API version detection
• Support for EEVEE Next rendering (Blender 4.2+)
• Backwards compatible with Blender 4.0-4.1

---

🤖 Generated with Claude Code

TexCut v2.2.0 - Simplified Dependencies

What's New in v2.2.0

🎉 Simplified Dependencies!

Shapely is no longer required - only OpenCV needed!

Removed

• Shapely dependency completely removed
• Removed fix_self_intersections() function
• Removed INSTALL_SHAPELY.md - no longer needed!

Changed

• Minimum boundary offset is now 1 pixel (was 0)
• Prevents self-intersections through dilation alone

Why This Change?

Comprehensive testing showed that boundary dilation alone prevents 100% of self-intersections when boundary_offset >= 1. The Shapely-based fix was redundant.

Benefits:

• ✅ Simpler Installation - One dependency instead of two
• ✅ Guaranteed Valid Polygons - 1px minimum dilation prevents issues
• ✅ Smaller Package - 9.7 KB (vs 11.6 KB)
• ✅ Fewer Failure Points - No Shapely installation issues

Installation

Step 1: Install OpenCV (One-Time)

OpenCV is not included with Blender by default. Install it once:

Windows:

cd "C:\Program Files\Blender Foundation\Blender 3.x\3.x\python\bin"
python.exe -m pip install opencv-python

macOS:

cd /Applications/Blender.app/Contents/Resources/3.x/python/bin
./python3.x -m pip install opencv-python

Linux:

cd /usr/share/blender/3.x/python/bin
./python3.x -m pip install opencv-python

Step 2: Install TexCut

1. Download TexCut-v2.2.0-20251107.zip
2. In Blender: Edit → Preferences → Add-ons → Install
3. Select the zip file and enable TexCut

Requirements

• NumPy ✅ (included with Blender)
• Pillow ✅ (included with Blender)
• OpenCV ⚠️ (requires installation - see above)

Test Results

See SELF_INTERSECTION_TEST_RESULTS.md for detailed analysis proving dilation alone is sufficient.

---

Full Changelog: https://github.com/rawnsley/TexCut/blob/main/CHANGELOG.md

TexCut v1.0.1 - Bug Fix Release

TexCut v1.0.1

Bug fix release addressing missing image file handling.

Fixed

• Fixed error when processing images whose source files have been moved or deleted
• Images with missing source files are now automatically saved to temporary location
• Better handling of packed and generated images

Changes

• Added file existence check before using image filepath
• Improved error handling for missing image files

Installation

1. Download TexCut-v1.0.1.zip from the assets below
2. In Blender, go to Edit > Preferences > Add-ons
3. Click Install... and select the downloaded zip file
4. Enable the add-on by checking the checkbox next to "Image: TexCut"

Usage

1. Open an image in Blender's Image Editor
2. Press N to open the sidebar
3. Go to the "Image" tab
4. Click "Create Mesh from Image"
5. Adjust settings as needed

Requirements

• Blender 3.0+
• Python packages: numpy, Pillow, opencv-python (see README for installation)

TexCut v1.0.0 - Initial Release

TexCut v1.0.0

First release of TexCut - a Blender add-on that creates optimized 2D meshes from images with transparency.

Features

• Works directly from the Image Editor
• Analyzes image alpha channel to detect non-transparent regions
• Generates mesh geometry that follows the transparency outline
• Automatically applies the image as a texture with proper alpha blending
• Adjustable alpha threshold and mesh simplification settings
• Maintains image aspect ratio

Installation

1. Download TexCut-v1.0.0.zip from the assets below
2. In Blender, go to Edit > Preferences > Add-ons
3. Click Install... and select the downloaded zip file
4. Enable the add-on by checking the checkbox next to "Image: TexCut"

Usage

1. Open an image in Blender's Image Editor
2. Press N to open the sidebar
3. Go to the "Image" tab
4. Click "Create Mesh from Image"
5. Adjust settings as needed

Requirements

• Blender 3.0+
• Python packages: numpy, Pillow, opencv-python (see README for installation)