Datasets:
| license: unknown | |
| tags: | |
| - microplastics | |
| - particle-classification | |
| - instance-segmentation | |
| - materials-science | |
| # mp-classification | |
| Masks and shape labels for microplastic particles detected in camera images | |
| of a microfluidic chip. Used to train a YOLO11-seg detector and a | |
| shape-based PS/PP/PET classifier. | |
| **339 instances, 241 images.** 257 instances are detection-training-eligible | |
| (`detection_gt_included=True`). 334 are labeled: **pp** 179 / **ps** 105 / | |
| **pet** 50. | |
| ## Layout | |
| ``` | |
| gt_instances.csv | |
| images/{source_image_id}.png (241 files) | |
| masks/{source_image_id}/particleN_mask.npz (339 files) | |
| ``` | |
| ## Related | |
| Paired trained models: | |
| [Inseop97/mp-segmentation-models](https://huggingface.co/Inseop97/mp-segmentation-models). | |
| ## License | |
| TBD. | |