Trucks Damage Dataset
This dataset contains images of trucks with various types of damage annotations. The annotations are in COCO format and include detailed segmentation masks for different types of damage.
Dataset Statistics
- Total Images: 795
- Annotations per category:
- scratch: 1125
- dent: 745
- broken part: 278
- missing part: 374
- crack: 87
- lamp broken: 6
Dataset Statistics
- Total Images: 795
- Annotations per category:
- scratch: 1125
- dent: 745
- broken part: 278
- missing part: 374
- crack: 87
- lamp broken: 6
Dataset Description
- Number of Images: 795
- Format: COCO
- Image Format: JPG
- Annotation Format: JSON
Damage Categories:
- Scratch
- Dent
- Missing Part
- Broken Part
- Flat Tire
- Crack
Annotation Format
Each annotation includes:
- Category ID
- Segmentation polygons
- Bounding box coordinates [x, y, width, height]
- Area
- Additional metadata
Usage
The dataset follows the COCO format and can be used with popular computer vision frameworks like:
- YOLO
- Detectron2
- MMDetection
- Transformers
License
Please specify the license under which this dataset is released.