滴度及注射方式对2型腺相关病毒载体表达效果的影响

由于腺相关病毒载体(adeno-associated viral vector,AAV)本身的侵染不具有神经元特异性,其在神经系统相关研究中会存在侵染范围不符合试验要求的情况,因此本试验拟研究不同滴度和注射方式对病毒侵染范围的影响。结果显示,在注射较高滴度的AAV2-CMV-EGFP(1.3×10¹¹mg/L)3周后,其侵染范围可由延髓中缝苍白核(raphe pallidus nucleus,RPa)区域延伸至下丘脑室旁核(paraventricular nucleus,PVN)区域;较低滴度(1.3×10⁸mg/L)结合压力注射的方式可以在一定程度上限制AAV的侵染范围,并且载体携带的EGFP荧光基团在637 nm波段不具有明显的假阳性信号。     

基于卷积神经网络的拖拉机工况识别

农机工况识别在细化农机作业状态和帮助掌握区域污染物排放趋势方面有着重要的研究价值。基于拖拉机不同运行状态下的行驶速度、发动机转速以及实时油耗等时间序列,首次提出将图像识别方法引入到拖拉机工况识别中的思路,并分别应用参数优化的支持向量机与卷积神经网络对实际作业拖拉机工况进行研究。结果表明:(1)基于参数优化的支持向量机可以较好地实现样本点的工况识别且识别准确度达到99.851 9%,但无法实现农机工况的连续性识别,同时无法对农机工况转换阶段进行有效识别。(2)以拖拉机运行速度与发动机转速等信息构建样本图像来描述农机工况变化的数据表达,并在此基础上应用卷积神经网络可以有效实现农机工况的连续性识别,且识别准确率可以达到93.3%。本研究在农机工况识别方面具有一定参考价值,并为后续农机不同工况下区域污染物排放研究提供技术支持。     

基于HSV与形状特征融合的花椒图像识别

光照条件变化会对花椒目标识别率产生影响,关系到机器视觉技术能否有效应用于生产现场的花椒采摘。通过对HSV特性图像识别技术的分析,提出HSV和形状特征融合的花椒识别算法。该算法采用同态滤波方法对光照进行补偿,解决因为光照不均匀而导致的花椒识别率低的问题,最后利用花椒圆度特征,排除树枝及树叶等的干扰,实现花椒的准确识别。利用同态滤波方法对光照进行补偿,对于光照不强或者发生遮挡的花椒图像有较大改善,通过试验得出其平均识别率达到94.0%,比单独采用HSV特性识别时,在顺光,背光和遮阴条件下,识别率分别提高4%,13%和21%,此外在遮阴条件下运行时间缩短14.6%。为遮阴条件下提高花椒识别率提供一种方法。     

Change features of time-series climate variables from 1962 to 2016 in Inner Mongolia,China

Detecting change features of climate variables in arid/semi-arid areas is essential for understanding related climate change patterns and the driving and evolution mechanism between climate and arid/semi-arid ecosystems.This paper takes Inner Mongolia of China,a unique arid/semi-arid ecosystem,as the study area.We first detected trend features of climate variables using the linear trend analysis method and then detected their trend-shift features using the breaks for additive seasonal and trend method based on the time-series of monthly precipitation and monthly mean temperature datasets from 1962 to 2016.We analyzed the different change features of precipitation and temperature on a regional scale and in different ecological zones to discover the spatial heterogeneity of change features.The results showed that Inner Mongolia has become warmer-wetter during the past 54 years.The regional annual mean temperature increased 0.4°C per decade with a change rate of 56.2%.The regional annual precipitation increased 0.07 mm per decade with a slightly change rate of about 1.7%,but the trend was not statistically significant.The warmer trend was contributed by the same positive trend in each season,while the wetter trend was contributed by the negative trend of the summer precipitation and the positive trend of the other three seasons.The regional monthly precipitation series had a trend-shift pattern with a structural breakpoint in the year 1999,while the regional monthly mean temperature series showed an increasing trend without a periodical trend-shift.After the year 2000,the warmer-wetter trend of the climate in Inner Mongolia was accelerated.The late 20th century was a key period,because the acceleration of the wetter trend in some local zones(I and II)and the alleviation of the warmer trend in some local zones(Ⅶ,Ⅷand IX)occurred simultaneously.Moreover,the change features had a strong spatial heterogeneity,the southeastern and southwestern of Inner Mongolia went through a warmer-drier trend compared with the other areas.The spatio-temporal heterogeneity of the climate change features is a necessary background for various types of research,such as regional climate change,the evolution of arid/semi-arid ecosystems,and the interaction mechanisms between climate and arid/semi-arid ecosystems based on earth-system models in Inner Mongolia.     

基于声振信号对称极坐标图像的苹果霉心病早期检测

为实现苹果早期霉心病较高精度的检测，该研究采用对称极坐标法（Symmetrized Dot Pattern，SDP）将苹果声振信号变换为雪花图，然后采用AlexNet、VGG16和ResNet50卷积神经网络以迁移学习方式深度挖掘SDP雪花图像的特征信息，将其输入到支持向量机（Support Vector Machine，SVM）分类器，对霉心程度≤7%的苹果进行检测。研究结果表明，当时间间隔系数为25和角度放大因子为50°时，健康果与早期霉心果声振信号的SDP图形状特征差异最大，在此条件下获取的SDP图经卷积神经网络AlexNet、VGG16和ResNet50提取特征并构建了不同核函数的SVM霉心果检测模型，在各类SVM模型中，ResNet50-SVM-gaus（高斯基）模型用相对较少的训练时间和参数量可取得训练集霉心果较高分类准确率，经超参数优化训练该模型对健康果和早期霉心果测试集不平衡样本（10∶1）的总体分类准确率达到96.97%，平均查准率、平均查全率、平均加权调和均值、Kappa系数和马修斯相关系数值分别为80.19%、90.36%、86.21%，82.54%和82.68%，该模型不仅对多数类的健康果保持较高分类准确率，而且对少数类的早期霉心果也具有较高判别能力。这些研究结果为声振法应用于果蔬内部病害的早期在线检测系统研发提供了技术支撑。     

Comparison of estimators and feature selection procedures in forest inventory based on airborne laser scanning and digital aerial imagery

Digital maps of forest resources are a crucial factor in successful forestry applications. Since manual measurement of this data on large areas is infeasible, maps must be constructed using a sample field data set and a prediction model constructed from remote sensing materials, of which airborne laser scanning (ALS) data and aerial images are currently widely used in management planning inventories. ALS data is suitable for the prediction of variables related to the size and volume of trees, whereas optical imagery helps in improving distinction between tree species. We studied the prediction of forest attributes using field data from National Forest Inventory complemented with ad hoc field plots in combination with ALS and aerial imagery data in Aland province, Finland. We applied feature selection with genetic algorithm and greedy forward selection and compared multiple linear and nonlinear estimators. Maximally around 40 features from a total of 154 were required to achieve the best prediction performances. Tree height was predicted with normalized root mean squared error value of 0.1 and tree volume with a value around 0.25. Predicting the volumes of spruce and broadleaved trees was the most challenging due to small proportions of these tree species in the study area.     

基于机器视觉的温室番茄裂果检测

该文通过对温室番茄果实进行定位及裂果检测,可为番茄裂果率预估及后续裂果自动筛选提供参考。针对自然光照下采集的各类番茄图像,在相关颜色空间中进行阈值预分割,利用前期支持向量机训练得到的纹理特征分类器对预分割区域进行二次判别;之后在前景区域利用显著性角点分割构造边缘轮廓集,利用基于最小二乘法修正的改进霍夫变换拟合单个番茄目标;最后利用二维Gabor小波算子对拟合的单个番茄区域进行纹理特征提取及裂果判别。文中共采集82幅番茄图像,其中50幅图像作为训练集图像,32幅图像作为验证集,所提算法对测试集中总共128个番茄的果实正确检出率为91.41%,对其中35裂果的正确判别率为97.14%,裂果判别部分平均耗时21 ms。试验结果表明,该方法具有较好的鲁棒性与可靠性,对成熟期番茄裂果率的估计研究及采摘过程中裂果的自动分级筛选具有较好的指导意义,为未来实现温室番茄果实生长状态在线监测提供参考。     

Identifying Soil and Plant Nutrition Factors Affecting Yield in Irrigated Mature Pistachio Orchards

The main objective of this study was to evaluate the potential use of a hybrid Genetic Algorithm-Artificial Neural Network (GA–ANN) method for predicting pistachio yield and for identifying the determinant factors affecting pistachio yield in Rafsanjan region, Iran. A total of 142 pistachio orchards were selected randomly and soil samples were taken at three depths. Besides, water samples and leaves from branches without fruit were taken in each sampling point. Management information and pistachio yields were achieved by completing a questionnaire. Primarily, 58 variables affecting pistachio yield were measured, and then 26 out of them were selected by minimizing mean square error (MSE) using a feature selection (FS) method. The results showed that the accuracy of the method was acceptable. Furthermore, the sensitivity analysis showed that the main determinant features affecting the pistachio yield were the irrigation water amount, leaf phosphorus, soil soluble magnesium, electrical conductivity (EC), and leaf nitrogen.     

基于变分模态分解-BA-LSSVM算法的配电网短期负荷预测

配电台区日负荷序列呈现为既包含变化趋势、又含有波动细节的不规则曲线,该文借助变分模态分解(variational mode decomposition,VMD)将包含这些信息的原始日负荷序列分解为不同频率尺度的子序列,并结合一系列复杂的环境因素,分别利用不同的最小二乘支持向量机(least squares support vector machine,LSSVM)模型进行负荷预测,最后将基于不同频率分量的预测结果相加得到最终的日负荷预测结果。为了提高LSSVM预测能力,采用蝙蝠算法(bat algorithm,BA)对各LSSVM的参数进行寻优,同时,该文分析了影响负荷变化的环境因素,设计了一套因素归一化方法,预测过程考虑了环境因素的影响。仿真结果表明,该文提出的考虑复杂环境因素的预测思想及对历史日负荷进行VMD分解、BA优化、LSSVM预测的组合预测方法能有效提高短期日负荷预测的准确性。     

基于太赫兹衰减全反射技术的花生霉变程度判别

为了能够可靠、快速、便捷地检测花生仁不同程度的霉变,研究了一种基于太赫兹时域光谱技术、分别结合误差反向传播(Back propagation,BP)神经网络算法与支持向量机算法(Support vector machine,SVM)的霉变花生定性分析方法。为排除不同样本带来的偶然性,实验随机采集花育36号、鲁花9号两个花生品种进行霉变培养。依据花生的感官特征与前人的研究经验,将花生分为正常、轻度霉变、中度霉变与严重霉变4类,采用太赫兹衰减全反射技术采集花生仁样本光谱(波段0. 3～3. 6 THz)。利用傅里叶变换方法对时域光谱信号进行频域变换并进行加窗处理,然后对所得频域信号进行光学常数吸光度与吸收系数的提取,得到样本的光学常数信号,并进行特征波段筛选。在此基础上分别建立BP神经网络定性分析模型与SVM定性分析模型。实验表明,BP神经网络模型对花育36号花生霉变模型的预测集识别正确率为88. 57%,对鲁花9号花生霉变模型的预测集识别正确率为91. 40%;Lib-SVM模型对两个品种花生霉变的二分类模型、3类霉变花生的三分类模型的预测集识别正确率均为100%。应用太赫兹时域光谱技术结合SVM算法检测霉变花生仁效果良好,具有一定的可行性。