Datasets:

kshitij1507
/

PCB-AOI

text string
0 0.79296875 0.846875 0.0265625 0.0359375
0 0.565625 0.234375 0.025 0.040625
0 0.57890625 0.52578125 0.0265625 0.0390625
0 0.88125 0.74609375 0.0296875 0.0390625
0 0.48515625 0.2265625 0.0296875 0.0421875
0 0.5828125 0.5015625 0.03125 0.0421875
0 0.821875 0.8203125 0.0234375 0.034375
0 0.5421875 0.2296875 0.021875 0.0390625
0 0.58046875 0.51875 0.0234375 0.0375
0 0.8578125 0.83828125 0.0421875 0.025
0 0.496875 0.78671875 0.0375 0.015625
0 0.478125 0.50625 0.0359375 0.0203125
0 0.8796875 0.1859375 0.0234375 0.0421875
0 0.784375 0.53828125 0.0140625 0.0359375
0 0.50390625 0.521875 0.01875 0.0359375
0 0.8625 0.16640625 0.021875 0.04140625
0 0.7859375 0.5234375 0.01328125 0.0359375
0 0.5046875 0.521875 0.0171875 0.034375
0 0.2875 0.1359375 0.0296875 0.01875
0 0.49140625 0.55 0.03125 0.0203125
0 0.36171875 0.3171875 0.0265625 0.01875
0 0.88125 0.3203125 0.01640625 0.028125
0 0.45078125 0.48671875 0.0171875 0.0296875
0 0.69453125 0.378125 0.0171875 0.02421875
0 0.67265625 0.884375 0.028125 0.0171875
0 0.5140625 0.45078125 0.0296875 0.0171875
0 0.61875 0.696875 0.025 0.0171875
0 0.1578125 0.5421875 0.0375 0.0265625
0 0.5859375 0.87109375 0.040625 0.025
0 0.75078125 0.86875 0.040625 0.0265625
0 0.15625 0.52421875 0.0375 0.025
0 0.56640625 0.87578125 0.0390625 0.021875
0 0.73125 0.88203125 0.040625 0.025
0 0.5421875 0.1578125 0.01796875 0.03359375
0 0.11875 0.49375 0.01640625 0.0375
0 0.08125 0.65390625 0.01953125 0.03671875
0 0.27421875 0.20859375 0.01875 0.0421875
0 0.10390625 0.21015625 0.02421875 0.05078125
0 0.35859375 0.6796875 0.0140625 0.03515625
0 0.5421875 0.26796875 0.02109375 0.04296875
0 0.2546875 0.225 0.021875 0.04375
0 0.084375 0.23828125 0.028125 0.0515625
0 0.37109375 0.6890625 0.0171875 0.0359375
0 0.52578125 0.265625 0.025 0.0453125
0 0.33359375 0.17109375 0.0265625 0.0453125
0 0.16640625 0.13984375 0.034375 0.053125
0 0.3265625 0.64921875 0.0203125 0.0359375
0 0.5859375 0.28046875 0.0296875 0.046875
0 0.4859375 0.321875 0.01875 0.0390625
0 0.446875 0.77890625 0.01953125 0.04140625
0 0.69375 0.52265625 0.0234375 0.0375
0 0.5296875 0.3234375 0.028125 0.0421875
0 0.38125 0.7578125 0.028125 0.04375
0 0.6828125 0.56875 0.03125 0.0421875
0 0.5015625 0.32109375 0.021875 0.040625
0 0.42265625 0.7734375 0.021875 0.0421875
0 0.69140625 0.53984375 0.0265625 0.0390625
0 0.58125 0.23046875 0.01875 0.034375
0 0.47109375 0.68125 0.0265625 0.05078125
0 0.8609375 0.1671875 0.02265625 0.03515625
0 0.5953125 0.23515625 0.0203125 0.0359375
0 0.46171875 0.6796875 0.0296875 0.0515625
0 0.878125 0.18671875 0.0234375 0.0359375
0 0.22890625 0.4234375 0.034375 0.01875
0 0.68203125 0.52578125 0.05 0.028125
0 0.1609375 0.1453125 0.034375 0.021875
0 0.91953125 0.4984375 0.021875 0.04375
0 0.84609375 0.89140625 0.025 0.0359375
0 0.14921875 0.69765625 0.021875 0.0390625
0 0.4578125 0.91796875 0.0453125 0.0265625
0 0.075 0.803125 0.0375 0.028125
0 0.33984375 0.13046875 0.040625 0.025
0 0.91796875 0.46171875 0.01953125 0.0421875
0 0.87890625 0.859375 0.02109375 0.034375
0 0.16796875 0.72734375 0.01875 0.03671875
0 0.40078125 0.57265625 0.046875 0.0296875
0 0.67578125 0.86015625 0.040625 0.0234375
0 0.2625 0.9078125 0.0390625 0.0234375
0 0.603125 0.4328125 0.046875 0.03125
0 0.34296875 0.13125 0.0421875 0.0265625
0 0.7578125 0.10546875 0.040625 0.025
0 0.4140625 0.41171875 0.0234375 0.04375
0 0.16796875 0.72421875 0.01875 0.0390625
0 0.06328125 0.321875 0.01953125 0.0375
0 0.54296875 0.26953125 0.01640625 0.0328125
0 0.11796875 0.49375 0.01875 0.040625
0 0.29296875 0.4109375 0.01640625 0.0375
0 0.728125 0.5546875 0.0328125 0.0171875
0 0.5265625 0.11875 0.0421875 0.0203125
0 0.6 0.29765625 0.0375 0.0171875
0 0.72890625 0.55078125 0.0328125 0.0171875
0 0.51953125 0.11875 0.040625 0.0203125
0 0.59609375 0.29609375 0.0375 0.0171875
0 0.4953125 0.70078125 0.046875 0.034375
0 0.51484375 0.5078125 0.04375 0.028125
0 0.5328125 0.16875 0.0515625 0.0375
0 0.6953125 0.45234375 0.02421875 0.0421875
0 0.50390625 0.48359375 0.01953125 0.040625
0 0.171875 0.55390625 0.0265625 0.04609375
0 0.5875 0.68125 0.046875 0.0328125

End of preview.

PCB-AOI Dataset

Overview

PCB-AOI is an object detection dataset for Automated Optical Inspection (AOI) of Printed Circuit Boards (PCBs). It contains annotated PCB images intended for training and evaluating models that detect defects, components, or hardware constraint violations on PCBs.

The dataset is structured in YOLO format, split into train, valid, and test sets, with a data.yaml configuration file describing the classes and paths.

Dataset Structure

PCB-AOI/
├── train/
│   ├── images/
│   └── labels/
├── valid/
│   ├── images/
│   └── labels/
├── test/
│   ├── images/
│   └── labels/
├── data.yaml
└── README.md

Each images/ folder contains PCB images, and each corresponding labels/ folder contains YOLO-format .txt annotation files (one per image, with normalized bounding box coordinates and class indices).

data.yaml

The data.yaml file defines the dataset configuration used by YOLO-based training frameworks (e.g., Ultralytics YOLO):

train: train/images
val: valid/images
test: test/images

nc: <number_of_classes>
names: ['class_1', 'class_2', '...']

Update the nc and names fields above to match the actual classes defined in your data.yaml.

Splits

Split	Description
`train`	Images and labels used for training
`valid`	Images and labels used for validation
`test`	Images and labels used for testing

Annotation Format

Annotations follow the standard YOLO format, where each label file contains one line per object:

<class_id> <x_center> <y_center> <width> <height>

All coordinates are normalized relative to the image width and height (values between 0 and 1).

Usage

Loading with Ultralytics YOLO

from ultralytics import YOLO

model = YOLO("yolov8n.pt")
model.train(data="data.yaml", epochs=100, imgsz=640)

Loading with Hugging Face `datasets` / `huggingface_hub`

from huggingface_hub import snapshot_download

dataset_path = snapshot_download(
    repo_id="kshitij1507/PCB-AOI",
    repo_type="dataset"
)

Intended Use

This dataset is intended for:

Training object detection models for PCB defect / component detection
Benchmarking AOI (Automated Optical Inspection) pipelines
Research on edge deployment of PCB inspection models (see the companion model repo: PCB-AOI-Hardware-Constraint)

Limitations

Dataset performance and generalization depend on lighting conditions, camera angle, and PCB type consistency with real-world deployment conditions.
Class imbalance across splits may affect model performance on underrepresented defect types.
Annotation quality has not been independently verified beyond the source labeling process.

License

Released under the MIT License.

Author

Kshitij Hedau

Hugging Face: kshitij1507

Citation

If you use this dataset in your work, please cite:

Kshitij Hedau.
PCB-AOI Dataset.
Hugging Face.

Downloads last month: 253

Total file size:

108 MB