Two postpartum cows were used to study the ultrastructural changes of uterine endometrium by using scanning electron microscope.The results showed that the process of repair of uterine endometrium after calving was de...Two postpartum cows were used to study the ultrastructural changes of uterine endometrium by using scanning electron microscope.The results showed that the process of repair of uterine endometrium after calving was demonstrated by scanning electron microscope through a series of endometrium biopsy.Some part of the endometrium was damaged after calving and its adjacent endometrium cells became necrosis and exfoliated during the first 7 days post-partum;the cilium and microvillus of the epithelial cell in the undamaged area of the endometrium disappeared.By 26 days postpartum the damaged area reduced and the cilium and microvillus increased in their numbers.The damaged tissues were all repaired by day 60 postpartum.展开更多
页岩扫描电镜(scanning electron microscope,SEM)图像智能识别能够快速分析页岩储层矿物,是页岩油储层“甜点”预测的重要手段之一,也是未来的技术发展趋势。传统方法在鉴定矿物成分时存在自动化程度低、样本适配度低和特征提取受限等...页岩扫描电镜(scanning electron microscope,SEM)图像智能识别能够快速分析页岩储层矿物,是页岩油储层“甜点”预测的重要手段之一,也是未来的技术发展趋势。传统方法在鉴定矿物成分时存在自动化程度低、样本适配度低和特征提取受限等问题。为此,本文提出基于BlendMask的页岩SEM图像鉴定方法。首先,采用双边滤波、拉普拉斯和图像归一化等图像预处理技术对原始图像进行去噪、锐化和像素统一处理,提高训练样本的质量;然后,采用旋转、缩放、光度变化等图像增广方法构建增广策略,扩大数据集数量;最后,利用注意力机制和深度可分离卷积改进BlendMask网络,实现图像的成分分割与识别。应用于海塔盆地的页岩SEM图像实验结果表明,相比BlendMask模型,改进后方法的分割准确率和召回率分别提升了0.02~0.20和0~0.59,分割用时减少了1.29~2.70 s。展开更多
基金The projectis supported financially by Heilongjiang( C- 970 2 ) and Chinese national( 39870 572 ) natural and scientificfunds.
文摘Two postpartum cows were used to study the ultrastructural changes of uterine endometrium by using scanning electron microscope.The results showed that the process of repair of uterine endometrium after calving was demonstrated by scanning electron microscope through a series of endometrium biopsy.Some part of the endometrium was damaged after calving and its adjacent endometrium cells became necrosis and exfoliated during the first 7 days post-partum;the cilium and microvillus of the epithelial cell in the undamaged area of the endometrium disappeared.By 26 days postpartum the damaged area reduced and the cilium and microvillus increased in their numbers.The damaged tissues were all repaired by day 60 postpartum.
文摘页岩扫描电镜(scanning electron microscope,SEM)图像智能识别能够快速分析页岩储层矿物,是页岩油储层“甜点”预测的重要手段之一,也是未来的技术发展趋势。传统方法在鉴定矿物成分时存在自动化程度低、样本适配度低和特征提取受限等问题。为此,本文提出基于BlendMask的页岩SEM图像鉴定方法。首先,采用双边滤波、拉普拉斯和图像归一化等图像预处理技术对原始图像进行去噪、锐化和像素统一处理,提高训练样本的质量;然后,采用旋转、缩放、光度变化等图像增广方法构建增广策略,扩大数据集数量;最后,利用注意力机制和深度可分离卷积改进BlendMask网络,实现图像的成分分割与识别。应用于海塔盆地的页岩SEM图像实验结果表明,相比BlendMask模型,改进后方法的分割准确率和召回率分别提升了0.02~0.20和0~0.59,分割用时减少了1.29~2.70 s。
