BLOOD CELL DETECTION AND CLASSIFICATION USING FASTER R-CNN, RETINANET, AND YOLOv12
Abstract
Detection and classification of blood cells play an important role in supporting the diagnosis of hematological diseases by reducing errors associated with manual microscopic examination. In this study, we propose an automated system for detecting and classifying 13 types of peripheral blood cells from microscopic images based on three deep learning models, namely RetinaNet, Faster R-CNN, and YOLOv12. Experimental results demonstrate that YOLOv12 achieves the best overall performance with stable detection capability and effective balance across different blood cell classes. These findings highlight the potential of deep learning–based object detection models for automated blood smear analysis and hematological diagnosis.
Tài liệu tham khảo
[1] World Health Organization, “Anaemia,” https://www.who.int/news-room/fact-sheets/detail/anaemia, (accessed 7 July 2026).
[2] Z. Liu, D. Yuan, and G. Zhu, “Automated Blood Cell Detection and Counting Based on Improved Object Detection Algorithm,” Mathematics, vol. 13, no. 18, 2025, doi: 10.3390/math13183023.
[3] H. Sazak and M. Kotan, “Automated Blood Cell Detection and Classification in Micro-scopic Images Using YOLOv11 and Optimized Weights,” Diagnostics, vol. 15, no. 1, 2025, doi: 10.3390/diagnostics15010022.
[4] X. Wang, G. Pan, Z. Hu, and A. Ge, “A two stage blood cell detection and classifica-tion algorithm based on improved YOLOv7 and EfficientNetv2,” Sci. Rep., vol. 15, no. 1, p. 8427, 2025, doi: 10.1038/s41598-025-91720-7.
[5] A. Abozeid et al., “White blood cells detection using deep learning in healthcare appli-cations,” Alexandria Engineering Journal, vol. 124, 2025, doi: 10.1016/j.aej.2025.03.098.
[6] Lorenzo Crumbie, “Hematopoiesis: Definition, embryology and cell lines,” https://www.kenhub.com/en/library/anatomy/histology-of-hematopoiesis, (accessed 15 May 2026).
[7] Julie Doll, “Cells of the blood,” https://www.kenhub.com/en/library/anatomy/cells-of-the-blood, (accessed 7 July 2026).
[8] T. Y. Lin, P. Goyal, R. Girshick, K. He, and P. Dollar, “Focal Loss for Dense Object Detection,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 42, no. 2, 2020, doi: 10.1109/TPAMI.2018.2858826.
[9] S. Ren, K. He, R. Girshick, and J. Sun, “Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks,” IEEE Trans. Pattern Anal. Mach. Intell., vol. 39, no. 6, 2017, doi: 10.1109/TPAMI.2016.2577031.
[10] Y. and Y. Q. and D. D. Tian, “Yolov12: Attention-centric real-time object detectors,” Adv. Neural Inf. Process. Syst., vol. 38, pp. 78433–78457, 2026.
[11] Sumith Singh Kothwal, “Blood Cell images for Cancer detection,” https://www.kaggle.com/dsv/10500753.
[12] M. Abou Ali, F. Dornaika, and I. Arganda-Carreras, “Blood Cell Revolution: Unveiling 11 Distinct Types with ‘Naturalize’ Augmentation,” Algorithms, vol. 16, no. 12, 2023, doi: 10.3390/a16120562.