The influence of NaCl and Na 2SO 4 on the germination rate and cotyledon outspread rate of Populus euphratica and P. pruinosa seeds from different seed sources was studied in this paper. The results showed that the se...The influence of NaCl and Na 2SO 4 on the germination rate and cotyledon outspread rate of Populus euphratica and P. pruinosa seeds from different seed sources was studied in this paper. The results showed that the seed relative germination rate and cotyledon outspread rate were negatively correlated with salt concentration significantly. The order of salt tolerance of P. euphratica seeds and P. pruinosa seeds from different seed sources was given respectively.In the same concentration,to the germination and cotyledon outspread rate of P. euphratica from different seed sources,NaCl was more influential than Na 2SO 4.In 0～0 4% concentration,to the relative seed germination rate and cotyledon outspread rate of P. pruinosa from different seed sources,NaCl was less influential than Na 2SO 4,and in 0 4%～1 5%, NaCl was more influential than Na 2SO 4 Compared with P. pruinosa, the salt tolerance of P. euphratica in seed germination period was better.展开更多
针对田间密植环境棉花精准打顶时,棉花顶芽因其小体积特性所带来识别困难问题,该研究提出一种改进型快速区域卷积神经网络(Faster Region Convolutional Neural Networks,Faster R-CNN)目标检测算法实现大田环境棉花顶芽识别。以Faster ...针对田间密植环境棉花精准打顶时,棉花顶芽因其小体积特性所带来识别困难问题,该研究提出一种改进型快速区域卷积神经网络(Faster Region Convolutional Neural Networks,Faster R-CNN)目标检测算法实现大田环境棉花顶芽识别。以Faster R-CNN为基础框架,使用RegNetX-6.4GF作为主干网络,以提高图像特征获取性能。将特征金字塔网络(Feature Pyramid Network,FPN)和导向锚框定位(Guided Anchoring,GA)机制相融合,实现锚框(Anchor)动态自适应生成。通过融合动态区域卷积神经网络(Dynamic Region Convolutional Neural Networks,Dynamic R-CNN),实现训练阶段检测模型自适应候选区域(Proposal)分布的动态变化。最后在目标候选区域(Region of Interest,ROI)中引入目标候选区域提取器(Generic ROI Extractor,GROIE)提高图像特征融合能力。采集自然环境下7种不同棉花总计4819张图片,建立微软常见物体图像识别库2017(Microsoft Common Objects in Context 2017,MS COCO 2017)格式的棉花顶芽图片数据集进行试验。结果表明,该研究提出方法的平均准确率均值(Mean Average Precision,MAP)为98.1%,模型的处理帧速(Frames Per Second,FPS)为10.3帧/s。其MAP在交并比(Intersection Over Union,IOU)为0.5时较Faster R-CNN、RetinaNet、Cascade R-CNN和RepPoints网络分别提高7.3%、78.9%、10.1%和8.3%。该研究算法在田间对于棉花顶芽识别具有较高的鲁棒性和精确度,为棉花精准打顶作业奠定基础。展开更多
文摘The influence of NaCl and Na 2SO 4 on the germination rate and cotyledon outspread rate of Populus euphratica and P. pruinosa seeds from different seed sources was studied in this paper. The results showed that the seed relative germination rate and cotyledon outspread rate were negatively correlated with salt concentration significantly. The order of salt tolerance of P. euphratica seeds and P. pruinosa seeds from different seed sources was given respectively.In the same concentration,to the germination and cotyledon outspread rate of P. euphratica from different seed sources,NaCl was more influential than Na 2SO 4.In 0～0 4% concentration,to the relative seed germination rate and cotyledon outspread rate of P. pruinosa from different seed sources,NaCl was less influential than Na 2SO 4,and in 0 4%～1 5%, NaCl was more influential than Na 2SO 4 Compared with P. pruinosa, the salt tolerance of P. euphratica in seed germination period was better.
文摘针对田间密植环境棉花精准打顶时,棉花顶芽因其小体积特性所带来识别困难问题,该研究提出一种改进型快速区域卷积神经网络(Faster Region Convolutional Neural Networks,Faster R-CNN)目标检测算法实现大田环境棉花顶芽识别。以Faster R-CNN为基础框架,使用RegNetX-6.4GF作为主干网络,以提高图像特征获取性能。将特征金字塔网络(Feature Pyramid Network,FPN)和导向锚框定位(Guided Anchoring,GA)机制相融合,实现锚框(Anchor)动态自适应生成。通过融合动态区域卷积神经网络(Dynamic Region Convolutional Neural Networks,Dynamic R-CNN),实现训练阶段检测模型自适应候选区域(Proposal)分布的动态变化。最后在目标候选区域(Region of Interest,ROI)中引入目标候选区域提取器(Generic ROI Extractor,GROIE)提高图像特征融合能力。采集自然环境下7种不同棉花总计4819张图片,建立微软常见物体图像识别库2017(Microsoft Common Objects in Context 2017,MS COCO 2017)格式的棉花顶芽图片数据集进行试验。结果表明,该研究提出方法的平均准确率均值(Mean Average Precision,MAP)为98.1%,模型的处理帧速(Frames Per Second,FPS)为10.3帧/s。其MAP在交并比(Intersection Over Union,IOU)为0.5时较Faster R-CNN、RetinaNet、Cascade R-CNN和RepPoints网络分别提高7.3%、78.9%、10.1%和8.3%。该研究算法在田间对于棉花顶芽识别具有较高的鲁棒性和精确度,为棉花精准打顶作业奠定基础。
