Scholarlink

We present a novel detection and classification of Plasmodium para site species using YOLOv11-based object detection system. Images of the malaria parasite species i.e., P. Falciparum, P. Malariae, P. Vivax, and P. Ovale are fine-tuned with pre-trained weights on MP-IDB dataset. Malaria which is a global health issue requires very accurate diagnosis for treatment and the control of disease. With the intention to balance the dataset to increase the performance of model, techniques of preprocessing which includes scaling, reduction of noise and mainly augmentation of data was done. Robust performance of detection was observed by the precision and recall of the model’ class. The training was done on the 80:20 dataset split into train and validation. The overall mAP@0.5 of 94.4% and mAP@0.5:0.95 of 75.2% was observed. The use of YOLOv11 engraves the path for enhanced dependable computer-aided system which can diagnose and help in precise treatment decision through viability and effectiveness.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Automated Detection and Classification of Multi-species Malaria Parasites Using YOLOv11

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Kanak Ohdar,
  • Shivang Priyadarshy,
  • Akriti Nigam

摘要

We present a novel detection and classification of Plasmodium para site species using YOLOv11-based object detection system. Images of the malaria parasite species i.e., P. Falciparum, P. Malariae, P. Vivax, and P. Ovale are fine-tuned with pre-trained weights on MP-IDB dataset. Malaria which is a global health issue requires very accurate diagnosis for treatment and the control of disease. With the intention to balance the dataset to increase the performance of model, techniques of preprocessing which includes scaling, reduction of noise and mainly augmentation of data was done. Robust performance of detection was observed by the precision and recall of the model’ class. The training was done on the 80:20 dataset split into train and validation. The overall mAP@0.5 of 94.4% and mAP@0.5:0.95 of 75.2% was observed. The use of YOLOv11 engraves the path for enhanced dependable computer-aided system which can diagnose and help in precise treatment decision through viability and effectiveness.