目前基于卷积神经网络将复合绝缘子喷水图像进行整体憎水性分类的方法对于图像局部的憎水性关注度不足,因此本文提出了一种基于目标检测算法的复合绝缘子表面憎水性判别方法。首先取样不同形态的喷水图片共5 800张,根据水珠形貌和接触角提出了单独针对水珠的分类标准。之后采用以SE-Resnet为骨架网络的Faster R-CNN对表面水珠进行分类,并获得了基于目标检测算法的21个水珠局部特征参数。为了兼顾图片全局特性,同时基于数字图像处理构建了12个与水珠亮斑面积和形态相关的全局参数。最后通过特征筛选 ...
何潇 +6 more
doaj
[Evaluation of brain injury caused by stick type blunt instruments based on convolutional neural network and finite element method]. [PDF]
Li H +8 more
europepmc +1 more source
[Sleep apnea automatic detection method based on convolutional neural network]. [PDF]
Gao Q, Shang L, Wu K.
europepmc +1 more source
[Research on fatigue recognition based on graph convolutional neural network and electroencephalogram signals]. [PDF]
Li S, Fu Y, Zhang Y, Lu G.
europepmc +1 more source
[Heart sound classification based on sub-band envelope and convolution neural network]. [PDF]
Wang X, Yang H, Zong R, Pan J, Wang W.
europepmc +1 more source
[An unsupervised three-dimensional medical image registration method based on shifted window Transformer and convolutional neural network]. [PDF]
Li Y, Ruan C, Ruan D.
europepmc +1 more source
[Research on the application of convolution neural network in the diagnosis of Alzheimer's disease]. [PDF]
Xu B, Ding C, Xu G.
europepmc +1 more source
基于深度卷积神经网络和迁移学习的农村房屋洪涝灾害后受损等级分类
Luyuan Wu +5 more
semanticscholar +1 more source
[Image reconstruction for cerebral hemorrhage based on improved densely-connected fully convolutional neural network]. [PDF]
Shi Y +5 more
europepmc +1 more source

