新疆典型盐渍区植被覆盖度遥感动态监测——以渭干河-库车河三角洲绿洲为例

以新疆典型盐渍区——渭干河-库车河三角洲绿洲为研究对象.首先,基于1989,2001,2006年的TM/ETM+遥感数据,以RS为技术手段,借助GIS空间分析方法、植被指数法,根据像元二分原理,采用复合植被指数VBSI估算植被覆盖度,其估算精度达到83.52%;其次,生成渭干河-库车河三角洲绿洲植被覆盖度分级图,并对其格局变化进行分析与评价,揭示出渭干河-库车河三角洲绿洲植被覆盖的时空变化规律.中高覆盖度、高覆盖度以林地、草地和耕地为主,对保持植被覆盖度起重要作用;林地面积的变化对重心转移起关键作用,而草地、耕地的波动则会影响到重心转移方向;林地、草地和耕地对于保持地区植被覆盖率有重要作用,应在长期土地利用规划中坚持植树造林,保持草地、保护耕地和控制未利用地;最后,利用植被覆盖度图像所提供的各覆盖度级的数量和空间分配状况来评价该绿洲植被覆盖度的动态变化,并进一步分析植被覆盖度变化的原因.基于遥感技术进行渭干河-库车河三角洲绿洲植被覆盖度动态变化分析的研究思路和方法能够为该绿洲生态环境保护提供参考依据.     

Qualitative and quantitative comparison of renal vs. hepatic ultrasonographic intensity in healthy dogs

A prospective, cross-sectional study was performed to qualitatively and quantitatively compare the echogenicity of the renal cortex relative to the liver in healthy dogs. Twenty-five normal adult dogs were examined ultrasonographically. Three standard B-mode images (8.0 MHz) and three tissue harmonic images of the cranial pole of the right kidney adjacent to the caudate lobe of the liver were obtained. Renal and hepatic echogenicities were qualitatively compared by two observers. Subsequently, regions of interest (ROIs) were drawn in the renal cortex and the adjacent liver parenchyma at equal depths on each image, using two different ROI geometries: deep adjacent half-annular ROIs centered at the focal zone and small superficial adjacent squares placed in the near field. Renal and hepatic mean pixel intensities were quantified and averaged for individual subjects. Qualitatively, the right renal cortex was more commonly hyperechoic to liver. Quantitatively, the renal cortical mean pixel intensity was significantly higher than that of liver using deep half-annular ROIs, but not superficial square ROIs, for both standard (P = 0.0007) and harmonic (P = 0.0107) tissue imaging. These findings suggest that the renal cortex can be slightly hyperechoic to adjacent liver. The framework within which the canine renal cortical parenchyma is routinely evaluated in abdominal ultrasonography should be reconsidered, and mild hyperechogenicity relative to the liver (at 8.0 MHz) interpreted as a normal finding.     

基于计算机视觉技术的砀山酥梨果梗识别

针对酥梨果梗计算机视觉检测方法的普适性问题,提出了一种利用像素点分析法识别果梗的算法。将酥梨图像进行二值化、滤波、形态学运算等预处理后,用Sobel算子提取酥梨图像边缘,再由一阶矩特征计算图像形心。选定形心为参考原点,建立图像的笛卡尔坐标系;从Y轴负方向开始,逆时针分别按15°和10°形心角对轮廓边缘点进行区域分块;统计分块区域内的像素点特征,并依此特征识别果梗。结果表明,对于无果梗的酥梨,15°和10°均能精确识别出来;有果梗的酥梨,15°较10°的识别率高,且含果梗区域的像素特征较其他区域的有很大突变。总体上果梗识别正确率可达97%,一定程度上可以满足酥梨果梗识别的需要。     

植物组织培养试管苗无菌检测技术

用图像处理技术对植物组织培养的试管苗进行图像处理,获得试管苗图像的像素点,同时也测定了试管苗的质量和体积。利用SPSS统计软件对像素点与质量、体积之间的函数关系进行曲线拟合,选择经检验达到显著水平的模型作为试管苗无菌检测模型。结果表明:用图像处理技术结合数学模型的应用,可以实现对试管苗的生长发育状况的无菌检测。     

Ovule dichotomy per locule of the trilocular pistil in cassava,Manihot esculenta: a useful abnormality for increased seed production?

Histologically, the cassava pistil is trilocular with a single ovule in each locule producing a maximum of three seeds in nature. Heat stress (35°/45°C, night/day temperatures), as a physical treatment was attempted to induce useful reproductive and vegetative traits in 10 F₁ hybrids derived from ‘OMR 36‐41’and ‘OMR 36‐42’crosses, and two Asian cassava cultivars,‘Rayong 60’and ‘Rayong 1′, under greenhouse and field conditions. Cytoembryological examination of the repro‐ ductive organs revealed morpho‐structural features of normal pistil in all clones except ‘Rayong 1’which exhibited reproductive variations under greenhouse culture. Genetically, an important variation was ovule dichotomy per locule of the trilocular pistil so that each pistil contained six instead of the usual three ovules and lacked any outer integument. These ovules possessed functional embryo sacs with varying nuclear number. A pair of the dichotomous ovules possessed two embryo sacs of apparent different megaspore mother cells. Vegetatively, F₁ hybrids were dwarf (0.50m) in their phenotype and flowered 1 month earlier than the 1.5‐2.0m tall cultivars under greenhouse conditions.     

A Method for Getting Directional Image from Fingerprint Based on Directional Elemental Pixel Sets

By the definition and interpretation of directional elemental pixel sets, a method for getting the directional image from the binary fingerprint image using the statistic characteristics of DEPS is presented in this paper. This method is better than traditional methods in processing speed, computing complexity and resisting noises. The directional image got by this method is used to the matching of ridges frame in a fingerprint verification system, and the testing result shows that the method is effective.     

基于机器视觉的田间小麦开花期判定方法

针对大量小麦育种材料花期难以精准、快速检测的问题,提出了一种基于综合颜色特征和超像素分割算法的小麦开花期判定方法。首先,根据光照强度及图像清晰度对综合颜色特征的过红颜色分量、HSV颜色空间的S分量和红绿归一化颜色分量自适应调节,增强小花和小穗的差异性。其次,基于中心距离函数和灰度变化函数改进超像素分割算法的聚类规则,获得由同质特征的相邻像素组成的图像区域。随后,优化图像区域路径搜索算法实现各图像区域精确分割,通过灰度和对比度指标完成各图像区域分类,实现小花与小穗的精准、快速分割,并根据小花与小穗的比例完成开花期判定。实验结果表明,本文所提出算法平均计算时间为0.172 s,小花平均识别精度为91%,小穗平均识别精度为90.9%,预测开花率与实际开花率的平均差值仅为1.16%,满足田间小麦开花期判定基本要求。     

基于像元二分模型的沙漠化地区植被覆盖度提取——以毛乌素沙地为例

以ETM＋为数据源，基于植被指数（NDVI）建立像元二分模型，对毛乌素沙地进行了植被覆盖度提取，然后根据实地调查数据对提取结果进行了精度验证，二者线性相关系数达到了0．92，平均精度为79．4％。研究结果表明，基于NDVI的像元二分模型适合于沙漠化地区的植被覆盖度提取。     

2001-2016年福建省海岸带植被覆盖变化特征

[目的] 揭示福建省海岸带地表植被覆盖空间格局变化，为开展基于主体功能区类型的资源环境监测与预警提供决策参考。[方法] 基于归一化植被指数（NDVI）像元二分模型，选用2001，2006，2011，2016年Landsat系列遥感数据，评估研究区植被覆盖度以表征评价区域生态承载力变化。[结果] ①2001-2016年福建省海岸带低植被覆盖度面积总体下降，较低和较高植被覆盖度变化较为平缓，而高植被覆盖度呈现上升趋势；②2001-2016年生态承载力状态改善的面积增加了10.78%，基本稳定面积变化不大，而退化面积下降了7.05%，显著退化面积下降了3.48%。③2001-2016年不同类型主体功能区内植被覆盖波动下降，并呈现以生态承载能力改善为主要变化特征。[结论] 2001-2016年福建省海岸带植被覆盖度和生态承载能力呈现总体上升态势，海岸带整体生态环境不断改善。     

震区植被覆盖度动态变化及其受损研究——以芦山县为例

本文探讨了利用Landsat TM和ETM影像数据建立植被受损模型,评价了植被受损的程度,运用植被覆盖度估算的像元二分模型,估算了植被覆盖度,总体上评价了芦山县经过4·20地震前后的植被覆盖度的变化情况,同时分析了其空间分布特征。研究结果显示:①2009年至2013年芦山县平均覆盖度从75.47%降低到70.22%②两个时期植被都是以较高覆盖度为主,约占总面积的36%,其中2009年高覆盖度植被面积大量的减少,减少的高覆盖度植被主要转化成了2013年中覆盖度的植被③2009年~2013年以轻度减少和稳定为主,轻度减少区和稳定区占总面积的75.66%,剧烈减少区和减少区占总面积的9.68%。这说明2009年~2013年植被有较严重的退化趋势④植被覆盖度变化存在着空间格局差异,随高程、坡度、坡向的分布特征差异显著⑤通过植被指数构建的植被受损模型评价了地震造成的植被受损情况,其中重度受损的植被大多是滑坡、崩塌的区域。通过研究植被受损的等级和覆盖度的动态变化情况,揭示了地震灾害对地表植被的破坏状况,为灾区的生态恢复提供决策依据。