一种基于数据增强与改进YOLO的结直肠镜息肉智能分类与检测方法
By using data augmentation and improved YOLO methods, realistic polyp-free images are generated. By combining the Swin Transformer with the improved YOLO model ST-YOLO, the problem of polyp detection in imbalanced datasets and complex backgrounds is solved, achieving high-precision polyp detection and classification.
CN121707930BActive Publication Date: 2026-07-17HARBIN INST OF TECH
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HARBIN INST OF TECH
- Filing Date
- 2025-11-24
- Publication Date
- 2026-07-17
Smart Images
Figure CN121707930B_ABST
Abstract
本发明公开了一种基于数据增强与改进YOLO的结直肠镜息肉智能分类与检测方法，所述方法如下：步骤1、对结直肠镜图像进行预处理与多样性增强；步骤2、对CycleGAN的训练损失函数进行改进，利用SC‑GAN模型生成无息肉健康结直肠镜图像；步骤3、在YOLOv8模型的主干网络Backbone层的空间金字塔池化模块前嵌入Swin Transformer模块得到ST‑YOLO模型；步骤4、将待检测图像输入至训练好的ST‑YOLO模型，判定有息肉或无息肉图像，然后进行差异化处理，对息肉目标进行进一步定位。本发明解决了现有技术中因数据不均衡及小目标检测性能不足所导致的息肉检测准确率低的问题。
Need to check novelty before this filing date? Find Prior Art