Blood group slide image dataset

Published: 9 September 2026| Version 2 | DOI: 10.17632/8m548j9m76.2
Contributors:
venkata sai charan koppuravuri,

Description

The dataset comprises 565 blood group images categorised into eight classes: A-positive, A-negative, B-positive, B-negative, AB-positive, AB-negative, O-positive, and O-negative. The original images are stored in the Original Dataset folder, while the corresponding bounding-box annotation files are provided in the Labels folder. The images were annotated using LabelImg, with bounding boxes generated for each individual blood drop corresponding to the Anti-A, Anti-B, Anti-D reagents. Each image is associated with its respective annotation file containing the coordinates of the annotated regions. All images in the dataset are provided in JPEG (.jpg) format.

Files

Institutions

Categories

Computer Vision, Annotation, Image Classification, Immunohematology, Deep Learning, Blood Analysis

Licence