Scholarlink

<p>To address the challenges of complex backgrounds and low detection accuracy in peony disease detection, this paper proposes an improved YOLO11n model, named YOLO11n-DSW. First, a Diverse Branch Block (DBB) module is introduced into the backbone network, utilizing multi-branch convolutional operations and contextual information fusion to enhance the feature extraction capabilities for peony disease targets of varying sizes. Second, a novel Scale-Converging Attention (SCA) mechanism is proposed, which effectively integrates features across different scales to detect detailed characteristics of peony disease regions. Additionally, the Wise-IoU (WIoU) loss function is adopted to replace the original Complete Intersection over Union (CIoU) loss function, better handling scale variations and shape differences, thus further improving detection accuracy. Experimental results show that the improved YOLO11n-DSW model achieves a detection accuracy of 84.8% in peony disease detection tasks, which is a 2.7% improvement over the original YOLO11n model. The mAP@0.5 and mAP@0.5:0.95 are 86.1% and 36.8%, respectively, representing improvements of 4.2% and 3%. Compared with other mainstream models, YOLO11n-DSW demonstrates superior detection performance in complex backgrounds, providing effective technical support for agricultural disease monitoring.</p>

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

The peony disease detection method based on the improved YOLO11n model

  • 期刊论文

  • 发表日期：

    2026年8月1日

查看全文
  • Junpeng Xu,
  • Jin Li,
  • Yu Wang,
  • Lei Shi

摘要

To address the challenges of complex backgrounds and low detection accuracy in peony disease detection, this paper proposes an improved YOLO11n model, named YOLO11n-DSW. First, a Diverse Branch Block (DBB) module is introduced into the backbone network, utilizing multi-branch convolutional operations and contextual information fusion to enhance the feature extraction capabilities for peony disease targets of varying sizes. Second, a novel Scale-Converging Attention (SCA) mechanism is proposed, which effectively integrates features across different scales to detect detailed characteristics of peony disease regions. Additionally, the Wise-IoU (WIoU) loss function is adopted to replace the original Complete Intersection over Union (CIoU) loss function, better handling scale variations and shape differences, thus further improving detection accuracy. Experimental results show that the improved YOLO11n-DSW model achieves a detection accuracy of 84.8% in peony disease detection tasks, which is a 2.7% improvement over the original YOLO11n model. The mAP@0.5 and mAP@0.5:0.95 are 86.1% and 36.8%, respectively, representing improvements of 4.2% and 3%. Compared with other mainstream models, YOLO11n-DSW demonstrates superior detection performance in complex backgrounds, providing effective technical support for agricultural disease monitoring.