在温室种植的粮食作物中检测高通量尖端,并使用基于深度学习模型的注意力机制
Sajid Ullah1,2,3, Klára Panzarová3, Martin Trtílek3
1Mendel Centre for Plant Genomics and Proteomics, Central European Institute of Technology (CEITEC), Masaryk University, Brno, Czech Republic.
Plant phenomics (Washington, D.C.)
|March 13, 2024
概括
这项研究通过改进的带有注意力机制的更快R-CNN (FRCNN-A) 模型,提高了作物中的谷物尖端检测. FRCNN-A实现了更高的准确性和更快的处理量化作物产量评估.
科学领域:
- 农业科学 农业科学
- 计算机视觉 计算机视觉
- 机器学习 机器学习
背景情况:
- 准确的作物产量评估依赖于检测谷物尖峰,这些尖峰通常被叶子掩盖.
- 现有的深度神经网络 (DNN) 由于它们的小尺寸和与树叶相似的外观,难以精确检测尖峰.
研究的目的:
- 为了提高图像中谷物尖峰检测的准确性和效率.
- 引入对更快R-CNN (FRCNN) 模型的架构修改,包括一个全球关注模块.
主要方法:
- 修改了Faster R-CNN架构,通过减少特征提取层和添加全球关注模块来创建FRCNN-A.
- 评估了FRCNN-A与传统的FRCNN和Swin Transformer对不同欧洲小麦品种的评估,包括具有挑战性的表型.
- 在基线和FastGAN增强数据集上测试模型,以在不同的数据条件下评估性能.
主要成果:
- 在FRCNN-A中,内部尖峰的检测精度得到了提高,平均精度为81.0%,而FRCNN的精度为76.0%.
- 在内尖上,FRCNN-A显示了与Swin变压器 (83.0% mAP) 相比具有竞争力的性能.
- 在数据集中,FRCNN-A被证明是一个比FRCNN和Swin Transformer更快,更轻的网络.
- 在增强数据上,FRCNN-A达到85.0% mAP,超过FRCNN (84.24%) 和接近Swin变压器 (89.45%).
结论:
- 建筑适应,特别是注意力机制,在检测微妙的谷物尖峰特征方面显著提高了DNN性能.
- FRCNN-A为基于图像的定量作物产量评估提供了一个有希望的,高效的解决方案.
相关概念视频
02:16Light Acquisition
8.5K
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.
8.5K
02:35Key Elements for Plant Nutrition
18.7K
Like all living organisms, plants require organic and inorganic nutrients to survive, reproduce, grow and maintain homeostasis. To identify nutrients that are essential for plant functioning, researchers have leveraged a technique called hydroponics. In hydroponic culture systems, plants are grown—without soil—in water-based solutions containing nutrients. At least 17 nutrients have been identified as essential elements required by plants. Plants acquire these elements from the...
18.7K