Fisher大间距线性分类器

作为一种著名的特征抽取方法,Fisher线性鉴别分析的基本思想是选择使得Fisher准则函数达到最大值的向量(称为最优鉴别向量)作为最优投影方向,以便使得高维输入空间中的模式样本在该向量投影后,在类间散度达到最大的同时,类内散度最小。大间距线性分类器是寻找一个最优投影矢量(最优分隔超平面的法向量),它可使得投影后的两类样本之间的分类间距(Margin)最大。为了获得更佳的识别效果,结合Fisher线性鉴别分析和大间距分类器的优点,提出了一种新的线性投影分类算法——Fisher大间距线性分类器。该分类器的主要思想就是寻找最优投影矢量wbest(最优超平面的法向量),使得高维输入空间中的样本模式在wbest上投影后,在使类间间距达到最大的同时,使类内离散度尽可能地小。并从理论上讨论了与其他线性分类器的联系。在ORL人脸库和FERET人脸数据库上的实验结果表明,该线性投影分类算法的识别率优于其他分类器。     

基于根点替代的猕猴桃果园行间导航线生成方法研究

针对棚架式猕猴桃果园树干分立、树干与树冠分离的生理特点,提出基于深度学习的目标检测框底边中点替代果树根点的导航特征目标检测方法。使用LabelImg标注预处理图像中的树干信息,生成猕猴桃树干数据集,基于Faster R-CNN构建目标检测模型,实现对猕猴桃果园行间有效范围的树干识别,利用根点替代方法确定树干根点的定位基点坐标,基于三次样条插值法提取树行线,根据最小二乘法拟合生成导航线。试验结果表明,生成的目标检测模型在地膜、杂草和土壤等果园行间环境下,树干正确识别率分别为90.6%、90.1%、89.4%;树干像素高度大于300像素、200～300像素、100～200像素的正确识别率分别为94.6%、89.9%、85.6%,表明选择树干作为果园行间视觉导航标识,有效利用了深度学习的稳定性,提高了导航目标识别的准确率;根点替代简化了导航定位基点生成算法,获取的果树定位基点与实际树干根点的平均横向像素误差占比为1.3%,平均纵向像素误差占比为2.2%,平均实际距离误差为0.102 m,导航定位基点精度较高;生成的导航线平均横向像素偏差为5.1像素,实际平均横向偏差为0.052 m,获取的果园行间导航路径满足行间导航要求。     

垄状微地形对近地表空气动力学特征的影响

在旱作农田中,耕作土垄常被用作风蚀控制系统的一部分,但其对近地表风速的影响机制尚不明确。基于此,文中利用风洞实验,测量了不同结构垄状微地形条件下的近地表风速。结果表明：气流通过垄覆盖区域时,在第一行垄周围风速变化最大,之后风速变化逐渐减小,趋于稳定,近地表风速沿风向表现为规律性的波动变化特征。相对于无垄状态,所有垄结构的风速降低率均在33%以上。垄结构不同,对近地表气流的影响机制不同。垄间距较小时,风速廓线形态以两段式为主,垄以粗糙元的形式影响气流;垄间距较大时,垄间风速减弱和恢复的水平距离增加,风速廓线形态以三段式为主,此时垄则起到风障的作用。     

基于遥感数据的新疆开-孔河流域农业区种植结构提取

农作物种植结构是农作物空间格局的重要组成部分,是区域土地资源和水资源优化配置的基础。以新疆开―孔河流域农业区为研究区域,综合利用作物物候信息和2016年的MODIS NDVI时序曲线,获得不同作物生长差异明显的关键期,选择关键期的Landsat 8 OLI影像,构建主要作物提取知识规则,基于决策树方法开展农作物的分类识别。开―孔河农业区2016年主要作物种植面积为5.07×10^5hm^2,其中棉花种植面积最大,为1.97×10^5hm^2,玉米、小麦次之。博斯腾湖和开都河农业区以辣椒、玉米和小麦为主要作物,种植结构比较零散;孔雀河农业区种植结构比较单一,以棉花和香梨为主要作物。与仅利用时间序列的MODIS数据进行作物分类识别的结果对比表明,综合利用MODIS和Landsat数据的作物识别精度有显著提高,总体分类精度从62.58%提高到88.37%,kappa系数从0.53提高到0.86。该方法综合利用了MODIS数据的时序特征和Landsat数据较高的空间分辨率特征,有效地避免了MODIS数据空间分辨率不足而导致分类精度较差的情况,又避免了Landsat数据时间分辨率不足所引起的时相选择盲目性或数据冗余,在干旱区农业种植结构的提取领域具有一定的应用价值。     

不同耕作方法对松嫩平原苏打盐碱化旱田改良利用效果试验

在最具有典型性和代表性的松嫩平原西部苏打盐碱土区(大安市)高梁田上,进行了不同深松耕法对苏打盐碱化旱田土壤容重、作物产量及相关性状的影响研究,结果表明:深松措施能够有效降低苏打盐碱化土壤容重,促进作物根系发育,显著提高作物(高梁)产量;深松措施可作为土壤紧实、容重较高的苏打盐碱土改良与高效利用的有效耕作方式加以实施。深松方法以效果极为显著且具有可操作性的处理1(垄沟深松40cm,中耕深松30cm)为佳,实施程序为原垄垄沟深松40cm—→撒施底肥—→破开原垄合成新垄—→中耕时垄沟深松30cm。     

高寒半干旱区马铃薯聚垄集肥覆膜技术效应

在河北张北小二台旱农试区研究表明，高寒半干旱区马铃薯聚垄集肥覆膜技术效应显著，生育前期具有明显的保水增温效果，且显著地促进了地上部和地下部生长发育和块茎与地上部比值的提高，但同聚垄覆膜处理相比，施肥在一定程度上又抑制了块茎与地上部比值的提高。同平种处理相比，聚垄集肥覆膜干薯产量提高185.3%～230．3％，作物水分利用效率提高10.9kg/(mm     

基于Sentinel-2影像的河套灌区作物种植结构提取

作物识别是提取作物种植结构的基础,利用遥感技术对作物进行监测识别,对优化生产布局、调整农业生产模式有着重要意义。文中选取河套灌区杭锦后旗为研究区域,基于2019年覆盖生长周期的Sentinel-2号卫星影像数据,构建NDVI时间序列数据集,利用Savitzky-Golay(S-G)滤波对NDVI时间序列数据集进行平滑,分析不同作物不同发育期的光谱曲线特征,计算各主要作物识别关键期的光谱阈值,构建基于决策树分层分类的农作物种植面积提取模型,并用验证样本对分类结果进行精度验证。结果表明:利用整个生育期内的NDVI最大合成影像确定植被地表覆盖,NDVI曲线变化区别林地与耕地,逐层提取地物,简便易行;采用S-G滤波重构高质量的NDVI时间序列曲线,研究证明重构后曲线更加平滑符合作物生长趋势;基于Sentinel-2号遥感数据和整个生育期NDVI时序数据,构建分层分类决策树模型,作物分类总体精度达92.1%,Kapppa系数精度达0.857。本研究采用的方法满足遥感观测应用化需求,也为县级区域农作物分类提供重要参考价值。     

沟垄集雨种植补灌对冬小麦旗叶叶绿素荧光参数的影响

为探明沟垄集雨种植技术在灌区对作物光合作用和产量的调控及节水效果,通过大田试验,研究了两种宽度的沟垄集雨种植结合不同补灌量处理对冬小麦光合速率、叶绿素荧光参数和产量的影响。结果表明:沟垄集雨处理下Pn较平作不灌水处理(CK)提高57.1%~84.4%,较平作灌水处理提高25.3%~50.3%;Fm、Fv、Fv/Fm、Fv/Fo、q P和籽粒产量也均高于平作处理,q N低于平作处理;两种宽度沟垄集雨种植较CK增产74.6%~124.9%,沟垄集雨补灌处理较平作灌水处理增产0.3%~18.9%,随灌水量增加产量增加,宽垄沟优于窄垄沟;在降雨较少的年份,60 cm沟垄集雨种植在灌溉量减少50%的条件下,光能转化效率比平作灌水处理提高44.1%,说明沟垄集雨种植可显著提高光能的转化效率及其量子产额,提高光合效率,维持较高的籽粒产量。     

大垄双行玉米免耕播种技术研究

主要针对我国东北冷凉风沙区在玉米播种时,难以实现原垄免耕播种的问题,提出了大垄双行玉米错茬免耕垄作播种的作业思路,大垄双行技术是指将传统的一垄一行改为一垄双行,垄台宽度约为传统垄台的3倍,目前设计为700 mm;建立了相应的试验田.试验表明:该项技术能够在田间大量玉米根茬直立覆盖情况下,在大垄上实现两行玉米错茬播种,保持垄形,作业质量满足农艺要求.与传统播种方式相比,具有提高地温、降低作业成本、提高产量等优点,可以提高地温约1℃,降低机械作业成本350元/hm2,提高产量14.42%,是保护性耕作技术在东北冷凉风沙区推广应用的一种较理想的模式.     

沟垄二元覆盖对旱地土壤水分及作物水分利用效率的影响

于2007年9月~2009年6月在渭北旱塬进行了冬小麦沟垄二元覆盖栽培模式的田间试验,研究了沟垄二元覆盖对旱地土壤水分及作物水分利用效率的影响。结果表明:垄覆地膜沟覆秸秆能显著提高产量,较对照(沟垄均不覆盖)产量最高增幅达39.62%(P0.05),垄覆地膜沟覆液膜处理两年平均产量较对照高32.16%,垄覆地膜沟不覆较对照高29.99%。垄覆地膜沟覆秸秆对小麦全生育时期0~40 cm土层的保墒作用明显,能有效提高土壤含水量,抽穗期0~40 cm土层含水量较对照高15.81%,沟覆液膜在小麦生育前期有良好的保水效果,返青期0~40 cm土层含水量较对照高11.01%。沟覆秸秆和沟覆液膜的水分利用效率较对照分别高31.11%(P0.05)和26.20%(P0.05)。     

一种基于冗余小波变换的DT网格运动估计和运动补偿方法

网格划分和网格顶点的运动估计是基于不规则网格的视频压缩技术的关键。为了进一步提高网格运动估计和运动补偿的效果,在综合比较现有冗余小波变换域运动估计方法和适用规则网格的EMRMC算法的基础上,提出了一种新的基于不规则网格的运动估计和运动补偿算法,即在冗余小波变换域提取特征点和运动潜在区,网格顶点的运动估计采用结合运动潜在区的在时域进行块匹配的运动估计和运动补偿方法,而运动补偿则通过三角形仿射变换完成。同时还给出了冗余小波变换域提取运动潜在区的计算模板。理论分析和实验结果表明,该算法在补偿效果方面较前两种方法得到了改进。     

Production and characterisation of monoclonal antibodies to cell wall components of the flax rust fungus

Monoclonal antibodies have been raised against an haustorium-enriched sample prepared from flax leaves infected with the biotrophic flax rust pathogen Melampsora lini. The monoclonal antibodies were produced following conventional and co-immunisation procedures and the range of antibody specificities was compared. The preparation used as immunogen for the conventional protocol was a crude isolate of haustoria consisting of approx. 65% fungal haustoria, the other components being mainly mesophyll cells or cell wall and chloroplast fragments. Following hybridoma production, 40% of positive cell lines produced antibodies that reacted with haustoria and other fungal cells, but 60% bound to plant cells in the infected leaves. For the co-immunisation protocol, the preparation used for immunisation consisted of the crude isolate of haustoria mixed with serum raised against an haustorium-depleted leaf homogenate. In two fusions, 92-94% of the antibodies reacted with fungal cells, including 3 cell lines that localised specifically to the cell wall of haustoria. Only 6-8% of the antibodies produced via co-immunisation reacted with plant cells. The antigens targeted by the three haustorium-specific monoclonal antibodies are incorporated into the wall at early stages of haustorium development, remain in the wall throughout haustorium maturation, and are present in both compatible and incompatible interactions. The epitopes recognised by the monoclonal antibodies are oligosaccharide in nature and the antigens are highly resistant to extraction from the wall. These results highlight the value of the co-immunisation protocol for the production of monoclonal antibodies to specific components in an impure preparation and provide direct evidence for molecular differentiation within the wall of the haustorium of M. lini.     

SAR图像边缘检测方法综述

边缘检测是SAR图像解译的基本问题之一。固有的乘性噪声特性使得经典的梯度边缘检测方法并不适用于SAR图像,因此研究针对SAR图像的边缘检测方法具有重要的意义。本文对SAR图像边缘检测问题进行了系统的研究,简单描述了SAR图像边缘检测问题,讨论了现有的SAR图像边缘检测方法,总结了SAR图像边缘检测中存在的两个问题——边缘细化和边缘定位,对检测结果的性能评估方法也进行了概括。最后在总结现有研究成果的基础上展望了SAR图像边缘检测的发展方向。     

酿酒高粱品种、组合及亲本的丝黑穗病抗性鉴定

为了明确不同类型的酿酒高粱材料对西南地区丝黑穗病的抗性和促进抗丝黑穗病品种的选育,采用土壤接菌法对107份材料进行了丝黑穗病抗性鉴定;根据抗性鉴定结果,比较了酿酒高粱杂交组合与其亲本对丝黑穗病的抗性,并对供试菌种进行了致病力分析.结果显示:在感病对照恢1植株平均病株率为67.65％时,有13个不育系、6个保持系、10个恢复系、28个杂交组合的病株率为0,达免疫水平,其中,酿酒高粱不育系和杂交组合所占比例较大,分别为86.67％和75.68％;恢复系中免疫类型所占比例较小,为24.39％;常规品种和组合中无免疫类型;杂交高粱对丝黑穗病的抗性属显性遗传;泸州丝黑穗病菌具有较强的致病力.     

Local convergence zones or discontinuous lines in the Taklimakan Desert, Northwest China

In order to study convergence zones or discontinuous lines formed locally in the Taklimakan Desert,we analyzed available MODIS images onboad TERRA or AQUA satellite for ten cases:29 March 2002,14 April 2002;2 January 2003;21 April 2003;26 March 2004;28 April 2004;30 January 2005;25 June 2005;3 December 2005;and 26 July 2006.We used weather maps published by the Japan Mete-orological Agency for the Asia and Pacific region at 500 hPa,700 hPa and 850 hPa,at 00Z and 12Z to analyze the meteorological conditions occurring in each case.We estimated the positions of air streams,convergence zones or discontinuous lines between them,and thermal conditions on the maps and con-firmed the presence of cumulus and cumulonimbus clouds through satellite images in particular.After a review of previous studies,this paper introduced the results of the present study.Closed warm areas in the Taklimakan Desert and on the Tibetan Plateau and local convergence zones or discontinuous lines in the Taklimakan Desert were discussed together with their diurnal changes,using composite maps of the ten cases.Along the long axis of the desert the convergence zones or discontinuous lines normally extend 70-80 km in a west-east direction,but are suspected to exceed 100 km in extreme cases.On the other hand,the convergence zones or discontinuous lines extending in a north-south direction on the southern fringe of the desert have a length of 40-60 km.The closed warm areas show clear diurnal changes,but they were not detected at the 500 hPa level.An example of a clear convergence zone running in a north-south direction on 26 July 2006 was presented in detail with corresponding satellite images.     

基于四元数旋转的开关型矢量中值滤波器

矢量中值滤波器是一种经典和高效的矢量滤波器,主要用于消除彩色图像中的冲击噪声。然而VMF没有区分细线条和噪声的能力,它往往把细线条当成噪声而过滤掉。本文利用四元数旋转理论,模仿Laplacian算子,提出一种用于检测彩色图像中的冲击噪声的算法,并结合传统的VMF构造出一个新颖的开关型矢量中值滤波器。实验结果表明,新的滤波器不仅能有效地保护细线条和边界等细节信息,而且其滤波性能也明显胜过传统的VMF和一些经典的及最近开发的矢量滤波器。     

Automatic sex detection of individuals of Ceratitis capitata by means of computer vision in a biofactory

BACKGROUND: The sterile insect technique (SIT) is acknowledged around the world as an effective method for biological pest control of Ceratitis capitata (Wiedemann). Sterile insects are produced in biofactories where one key issue is the selection of the progenitors that have to transmit specific genetic characteristics. Recombinant individuals must be removed as this colony is renewed. Nowadays, this task is performed manually, in a process that is extremely slow, painstaking and labour intensive, in which the sex of individuals must be identified. The paper explores the possibility of using vision sensors and pattern recognition algorithms for automated detection of recombinants. RESULTS: An automatic system is proposed and tested to inspect individual specimens of C. capitata using machine vision. It includes a backlighting system and image processing algorithms for determining the sex of live flies in five high-resolution images of each insect. The system is capable of identifying the sex of the flies by means of a program that analyses the contour of the abdomen, using fast Fourier transform features, to detect the presence of the ovipositor. Moreover, it can find the characteristic spatulate setae of males. Simulation tests with 1000 insects (5000 images) had 100% success in identifying male flies, with an error rate of 0.6% for female flies. CONCLUSION: This work establishes the basis for building a machine for the automatic detection and removal of recombinant individuals in the selection of progenitors for biofactories, which would have huge benefits for SIT around the globe.     

玉米圆斑病防治研究

目前吉林省推广的玉米骨干自交系吉63果穗上发生的穗腐症状是由Hel-minthosporium carbonum Ullstrup侵染引起的圆斑病。从病斑形态和培养性状与小斑病菌迥然不同。经温室苗期和田间成株期人工接种证明,圆斑病菌除为害吉63和少数自交系外,不侵染T型、M型等不育系及多数正常系玉米的果穗。从果穗抽出到灌浆期是圆斑病菌侵染发病的关键时期。病菌主要在病株残体里以菌丝体越冬,已腐烂的病组织则丧失生活力。六年共鉴定655份自交系、品种及杂交种,有184份材料的果穗感病,以吉63及其选系后代发病最重。经测定吉63感染圆斑病是由隐性感病基因所控制。     

New genes for disease resistances through somatic hybridization

Somatic hybridization, a process of combining protoplasts from different plants, can provide new sources of disease resistances for plants. In the case of wild and cultivatedSolanum species, the hybrids express resistances from each partner in the fusion and can often be crossed with cultivars to improve agronomic characteristics of the tubers. Restriction fragment length polymorphism (RFLP) analyses can provide a means for determining that the plants being investigated are actually hybrids as well as a means for following the introgression of DNA into progeny lines. These points are addressed in this paper with specific reference to somatic hybrids betweenSolanum brevidens and potato.