RSG-YOLOV8:基于增强的YOLOv8和多尺度注意力功能融合的芽率的检测
Huikang Li1, Longbao Liu1, Qi Li1
1College of Engineering, Anhui Agricultural University, Hefei, China.
PloS one
|November 12, 2024
概括
一个新的模型,米种发芽-YOLOV8 (RSG-YOLOV8),改善了米种发芽率的检测. 它通过集成CSPDenseNet等高级功能和一种新的注意力机制来提高准确性,从而设定了一个新的基准.
科学领域:
- 农业工程 农业工程
- 计算机视觉 计算机视觉
- 机器学习 机器学习
背景情况:
- 准确检测大米种子发芽率对于农业至关重要.
- 现有的方法在发芽和非发芽种子之间存在微妙的视觉差异,导致精度低.
研究的目的:
- 开发一个高度准确的深度学习模型来检测大米种子发芽率.
- 解决目前在分辨发芽和非发芽大米种子方面的方法的局限性.
主要方法:
- 拟议的米种子发芽-YOLOV8 (RSG-YOLOV8) 模型包含CSPDenseNet进行高效的计算.
- 集成了一个动态和稀疏的注意力机制 (BRA),以专注于关键特征.
- 采用基于GFPN的结构化特征融合网络来增强特征融合.
- 引入了一个额外的检测头,具有可变的箱尺度和优化的回归损失.
主要成果:
- RSG-YOLOV8在超过欧盟50% (mAP50) 门的交叉点实现了0.981.98的平均平均精度.
- 与标准YOLOv8模型相比,在mAP50中显示了4%的增强.
- 在大米种子发芽数据集上建立了一个新的基准性能.
结论:
- RSG-YOLOV8模型显著提高了大米种子发芽率检测的准确性.
- 结合CSPDenseNet,BRA关注,基于GFPN的融合和增强的检测头是有效的.
- 这项研究为自动化农业监测和质量控制提供了强大的解决方案.
相关概念视频
02:16Light Acquisition
8.4K
In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.
8.4K
03:21RNA-seq
9.8K
RNA sequencing, or RNA-Seq, is a high-throughput sequencing technology used to study the transcriptome of a cell. Transcriptomics helps to interpret the functional elements of a genome and identify the molecular constituents of an organism. Additionally, it also helps in understanding the development of an organism and the occurrence of diseases.
Before the discovery of RNA-seq, microarray-based methods and Sanger sequencing were used for transcriptome analysis. However, while...
Before the discovery of RNA-seq, microarray-based methods and Sanger sequencing were used for transcriptome analysis. However, while...
9.8K