femartip/Detection_Microfibers_APP
Application for object detection of microfibers in sludge samples extracted from residual water treatment plants.
What's novel
Application for object detection of microfibers in sludge samples extracted from residual water treatment plants.
Code Analysis
12 files read · 4 roundsA PySide6 desktop GUI application that loads pre-trained TorchScript object detection models to identify microfibers in sludge filter images, then measures fiber length via skeletonization and approximates fiber color via K-means clustering.
Strengths
The core detection pipeline is functional and covers a real domain need (microplastic detection in wastewater), with a reasonable post-processing chain (thresholding, mask fitting, skeletonization for length, K-means for color). The GUI provides a usable workflow for batch image processing with progress feedback.
Weaknesses
No tests exist, the requirements.txt is a 200+ package Anaconda dump making installation impractical, the 'CUDA' and 'CPU' model files are byte-identical (same SHA256), the box-merging logic is commented out/disabled, and the code mixes GUI concerns with processing logic in a monolithic class. The merge_boxes_and_masks function is effectively a no-op since the actual merging code is commented out.
Score Breakdown
Signal breakdown
Innovation
Craft
Traction
Scope
Evidence
Commits
8
Contributors
2
Files
313
Active weeks
5
Repository
Language
Python
Stars
3
Forks
1
License
MIT