全细胞脂肪酸气相色谱法鉴别莫拉氏菌和军团杆菌

本文用毛细管气相色谱法绘制了36株莫拉氏菌(Moraxella)及其类属菌和13株嗜肺军团杆菌(Legionella pneumophila)全细胞脂肪酸气相色谱图,用气相色谱-质谱法解析了某些色谱峰,借助电子计算机对脂肪酸色谱图进行聚类分析。结果表明,我国分离的莫拉氏菌新种(M.amylolytica和M.nanchang)和目前该菌主要标准菌株均不相同,脂肪酸成分以异十五烷酸为主。我国新分离的嗜肺军团菌N J8331和LDB 4a的脂肪酸色谱图和国际标准株基本一致,主要成分为异十六烷酸,而另一株LDB 4b和Legionella pneumophila 1-8型差别很大。以上结果和传统分类鉴定结果一致,由此看到化学分类法在微生物分类鉴定和生物学研究中的重要性。     

DISCRIMINATION OF Moraxella AND Legionella pneumophila BY GAS CHROMATOGRAPHY OF WHOLE-CELL FATTY ACIDS

Whole-cell fatty acid profiles of 36 strains of Moraxella and its related bacteria, and 13 strainsof Legionella pneumophila were obtained by capillary gas chromatography (GC). Some component peaksin these profiles were analyzed by gas chromatography-mass spectrometry (GC-MS). Hierarchical clus-tering analysis of these bacterial fatty acid patterns was done with a microcomputer. Results showedthat these were some obvious differences between the two new species of Moraxella isolated in China,i.e. M. amylolytica and M. nanchang, and the currently established reference strains of this genus. Themajor fatty acids of these two new species were the isomers of pentadecanoic (C_(15:0)) acids. Theprofiles of L. pneumophila NJ 8331 and LDB 4_a recently isolated in China were essentially similar tothose of international reference strains type 1-8 (CDC, USA), which contained mainly 16-carbon (C_(16))acids, whereas another isolate LDB 4b was obviously dissimilar. The results of this method of discrim-ination and of t     

基于视网膜大细胞通路计算模型和深度学习的无人机目标检测

在复杂背景下的小型无人机红外目标检测是计算机视觉领域的挑战性课题.传统目标检测算法利用深度卷积神经网络提取无人机的静态外观特征并进行模式判别,但在复杂背景下且目标外观不清晰时的性能会显著下降.本文借鉴生物视网膜机制,通过视网膜大细胞通路模型提取无人机目标的时空运动信息,同时借助深度卷积神经网络获得基于静态表观特征的目标置信度图,进而将视网膜时空运动信息与深度卷积网络的目标置信度图进行融合获得目标检测结果.在Anti-UAV2020公开数据集上的评估结果表明,所提出算法的检测精确率达到86.90%,超过了业内主流的YOLO-v3算法.