nodes/Limbicnation-ComfyUI-TransparencyBackgroundRemover
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
38 Commits 1 Branch 0 Tags
d49b2eea3255d38b0db3d1fd0e30ec27e4ff1ff4
Commit Graph
38 Commits
This Branch
This Branch
All Branches
Author SHA1 Message Date
limbicnation d49b2eea32 Bumping version 1.0.4 2025-08-21 22:06:18 +02:00
limbicnationandClaude 28c44c5184 feat(grabcut): add RGBA/MASK output format selection 
Co-Authored-By: Claude <noreply@anthropic.com>
 2025-08-21 21:45:38 +02:00
limbicnation b727939b99 fix(grabcut): resolve ComfyUI tensor format compatibility errors 
Fixes "Cannot handle this data type: (1, 1, 1024), |u1" error in PreviewImage node:

- Maintain (H,W,C) tensor format instead of transposing to (C,H,W)
- Add explicit float32 data type conversion for PyTorch tensors
- Implement comprehensive error handling with fallback values
- Fix OpenCV guided filter data type compatibility issues

Ensures GrabCut nodes output tensors in ComfyUI's expected format:
(batch, height, width, channels) for images and (batch, height, width) for masks.
 2025-08-21 20:46:17 +02:00
Gero Doll 98c23c27d8 Merge pull request #5 from Limbicnation/feat/content-aware-edge-detection 
Feat/content aware edge detection
 2025-08-19 02:14:49 +02:00
limbicnation 22098d9f06 fix: resolve critical runtime errors in edge detection 
- Fix IndexError in _combine_edge_maps() when edge_maps list is empty
- Fix ZeroDivisionError in _enhanced_sobel_edge_detection() for solid color images
- Fix NaN issue in _detect_pixel_art_characteristics() when no edges detected
- Make AUTO mode truly automatic by using content detection instead of dither_handling
 2025-08-19 02:07:21 +02:00
limbicnation 3d51cabb92 fix(edge-detection): resolve critical runtime errors and improve stability 
Fixed IndexError in combine_edge_maps when input is empty
Prevented ZeroDivisionError in enhanced_sobel_edge_detection for solid colors
Fixed NaN output in detect_pixel_art_characteristics when no edges found
Implemented true automatic mode by integrating content detection logic
 2025-08-19 01:30:07 +02:00
limbicnation 0a2171a3ff Update pyproject.toml with new dependencies and version 2025-08-19 01:21:03 +02:00
limbicnation edfc53f6b1 feat(edge-detection): Implement content-aware processing modes 
Adds a multi-method edge detection algorithm that automatically adapts to image content (Pixel Art vs. Photographic).

- Combines Roberts Cross, Sobel, and Canny methods for improved accuracy.

- Introduces an `edge_detection_mode` parameter (`AUTO`, `PIXEL_ART`, `PHOTOGRAPHIC`) for manual control.

- Optimizes processing to preserve sharp details in pixel art and create clean edges in photos.
 2025-08-19 01:16:30 +02:00
limbicnation 97f5833d13 feat: optimize edge detection for pixel art characters 
- Add multi-method edge detection (Roberts Cross, Enhanced Sobel, Pixel-aware Canny)
  - Implement automatic content type detection (pixel art vs photographic)
  - Add content-aware edge refinement with pixel-perfect preservation
  - Add new edge_detection_mode parameter (AUTO/PIXEL_ART/PHOTOGRAPHIC)
  - Optimize performance with scale-adaptive processing
  - Preserve sharp boundaries for pixel art while smoothing photographic content
 2025-08-19 00:07:56 +02:00
limbicnation 65e38085e0 Bump version 1.0.2 2025-06-04 02:28:34 +02:00
Gero Doll 0dd5b72602 Merge pull request #4 from Limbicnation/feature/enhance-core-functionality 
Add batch processing and performance
 2025-06-04 02:21:18 +02:00
limbicnation 9f86bc818f Fix CI imports and improve batch node docs 2025-06-04 02:09:05 +02:00
limbicnation 24df466093 Add batch processing and performance 2025-06-03 23:53:09 +02:00
Gero Doll adf22dc001 Merge pull request #3 from Limbicnation/fix/github-actions-pipeline 
Fix/GitHub actions pipeline
 2025-05-26 05:25:36 +02:00
limbicnation ad7781c252 Bump version to 1.0.1 2025-05-26 05:20:26 +02:00
limbicnation d97f948d68 Fix GitHub Actions pipeline issues 
- Update CI workflow with better dependency handling and validation
- Fix publish workflow token handling with fallback method
- Add missing scikit-learn dependency to requirements.txt
- Enhance error handling and validation steps
 2025-05-26 05:02:27 +02:00
limbicnation e2023bd772 Fix: Update repository name in toml file 2025-05-25 01:42:24 +02:00
limbicnation a6d15a166b Fix: Decapititize PublisherId in toml file 2025-05-25 01:32:19 +02:00
limbicnation a106dc49dc Fix: path for third demo image 2025-05-24 06:11:26 +02:00
limbicnation f6ddf18cca Update README visual demo (ComfyUI) 2025-05-24 06:09:30 +02:00
limbicnation 875600a3f8 Add second example image to README visual demo 2025-05-24 06:03:07 +02:00
limbicnation bedc28ea86 Create comprehensive README with installation guide and parameter documentation 2025-05-24 05:47:13 +02:00
Gero Doll 639ed2eed1 Merge pull request #2 from Limbicnation/feature/power-of-8-scaling 
Add power-of-8 scaling with intelligent NEAREST neighbor interpolation
 2025-05-24 04:34:30 +02:00
limbicnation eb9e21a48c Add power-of-8 scaling with intelligent NEAREST neighbor interpolation 
- Replace scale_factor with user-friendly output_size dropdown
- Support predefined dimensions: 64x64 through 2048x2048 (all power-of-8)
- Add intelligent scaling calculation with aspect ratio preservation
- Implement minimum 64x64 input validation
- Maintain pixel-perfect quality with NEAREST interpolation
 2025-05-24 04:17:48 +02:00
Gero Doll 674f6b13ec Merge pull request #1 from Limbicnation/fix/parameter-validation 
Fix/parameter validation
 2025-05-23 05:40:06 +02:00
limbicnation 7a10e85c85 Clean up debug prints and unused variables 2025-05-23 05:30:54 +02:00
limbicnation 358f720bbf Add debug prints and fix mask combination threshold 2025-05-23 04:27:20 +02:00
limbicnation a9c146966e Fix alpha channel: invert mask for proper transparency 2025-05-23 04:18:16 +02:00
limbicnation 9a9e4fd9bb Move binary_threshold to required params to fix validation 2025-05-23 04:10:36 +02:00
limbicnation 7dcf358fe5 Add default value to output_format enum parameter 2025-05-23 04:04:24 +02:00
limbicnation f15e6fcc58 Fix parameter order: binary_threshold before output_format 2025-05-23 04:01:39 +02:00
limbicnation f0ae1bd9c0 Fix parameter order for binary_threshold in ComfyUI node 2025-05-23 03:58:05 +02:00
limbicnation 97614a54cd Fix alpha mask greytones by adding binary thresholding 2025-05-23 03:45:13 +02:00
limbicnation d7162a1942 Add error handling and publish workflow 2025-05-23 02:03:45 +02:00
limbicnation 218fd597d0 Add GitHub CI workflow 2025-05-23 01:37:08 +02:00
limbicnation 7233d89f0a Add project config and workflow examples 2025-05-23 01:35:09 +02:00
limbicnation 933cf313fb Add ComfyUI transparency background remover nodes 2025-05-23 01:30:31 +02:00
Gero Doll 41ee4e2bcb Initial commit 2025-05-22 22:33:04 +02:00