Hyperspectral waveband selection for internal defect detection of pickling cucumbers and whole pickles

Hyperspectral imaging under transmittance mode has shown potential for detecting internal defect, however, the technique still cannot meet the online speed requirement because of the need to acquire and analyze a large amount of image data. This study was carried out to select important wavebands for further development of an online inspection system to detect internal defect in pickling cucumbers and whole pickles. Hyperspectral transmittance/reflectance images were acquired from normal and defective cucumbers and whole pickles using a prototype hyperspectral reflectance (400-740 nm)/transmittance (740-1000 nm) imaging system. Up to four-waveband subsets were determined by a branch and bound algorithm combined with the k-nearest neighbor classifier. Different waveband binning operations were also compared to determine the bandwidth requirement for each waveband combination. The highest classification accuracies of 94.7 and 82.9% were achieved using the optimal four-waveband sets of 745, 805, 965, and 985 nm at 20 nm spectral resolution for cucumbers and of 745, 765, 885, and 965 nm at 40 nm spectral resolution for whole pickles, respectively. The selected waveband sets will be useful for online quality detection of pickling cucumbers and pickles.     

Diagnosis of bacterial spot of tomato using spectral signatures

Ultraviolet, visible, and near-infrared reflectance spectroscopy was used to determine the disease severity of tomato (Lycopersicon esculentum) leaves infected with Xanthomonas perforans, the causal agent of bacterial leaf spot of tomato. Chemometric methods were used to identify significant wavelengths and create spectral-based prediction models. Significant wavelengths were identified through analysis of the B-matrix from partial least squares (PLS) regression, analysis of a correlation coefficient spectrum, and through the use of a stepwise multiple linear regression (SMLR) procedure. These analysis methods revealed several significant regions wavelengths and produced predictive models of disease severity based on absorbance spectra. The best model predicted the disease severity of the validation data set with a root mean square difference (RMSD) of 4.9% and a coefficient of determination (R²) of 0.82. The results of this initial study indicate the potential for the use of spectral technology to detect bacterial leaf spot of tomato in the field.     

玉米自交系品质性状的近红外光谱测定

利用近红外透射光谱分析仪,对黑龙江省农业科学院草业所玉米室选育的71份玉米自交系进行了品质成份测定,其中2005年的14份和2006年的57份。结果表明：玉米籽粒的3种成份在不同材料中差异显著,粗脂肪的变异系数较大,2005和2006年分别为7．09和7．46;粗淀粉的变异系数较小,为0．51和0．68;蛋白质的变异系数分别为4．03和4．51。自交系的粗脂肪平均含量分别为3．81％和3．62％,粗淀粉平均含量为72．44％和71．96％,蛋白质平均含量为12．15％和11．54％。同时针对近红外技术进行了讨论。     

Spectral measurements of the total aerial N and biomass dry weight in maize using a quadrilateral-view optic

Heterogeneous crop stands require locally adapted nitrogen fertilizer application based on rapid and precise measurements of the local crop nitrogen status. In the present study, we validated a promising technique for the latter, namely a tractor-mounted field spectrometer with an oblique quadrilateral-view measuring optic, measuring solar radiation and canopy reflectance in four directions simultaneously. Dry matter yield (kg ha⁻¹), total N content (g N g⁻¹ dry matter) and total aerial N (aboveground N-uptake) (kg N ha⁻¹) in maize were determined in 10 m² calibration areas in 60 plots differing in their N treatment and seeding density three times in each of three years under field conditions. Results show that the sensor used can reliably determine total aerial N ranging from as little as 5 kg N to 150 kg N ha⁻¹ with R²-values ≥0.81 in 2002 and 2004, and with R²-values ranging from ≥0.57 to 0.84 in 2003. Dry matter yields from as low as 0.3–4.2 t ha⁻¹ could be determined with R²-values ranging from 0.67 to 0.91 in 2002 to 2004. The capacity to ascertain DM yield spectrally was drastically reduced in the higher yield range (>6 t ha⁻¹) probably due to decreased sensitivity of the spectral signal. N-contents were generally not well determined. Taken together there is a good potential to determine reliably differences in total aerial N or DM yield from the five leaf stages unfolded to the five node stage where typically nitrogen applications are carried out.     

棉花群体结构与棉田光量子传递特性关系的研究

棉花群体结构与棉田光量子的透射、 反射和吸收等传递特性有密切关系, 其中透射率与叶面积指数和群体叶片重关系最密切, 可用于估测叶面积指数和群体叶重的大小。 光的透射率与群体地上部分生物量和茎枝生物量也有很好的关系, 可用于活体农作物生物量间接测定。 反射率与叶面积和生物量也有一定关系, 采用反射遥感方法也可在     

应用遥感技术监测土壤侵蚀动态的研究

本文阐述了应用遥感技术进行土壤侵蚀动态监测研究的机理，建立了我国南方地区土壤侵蚀程度解译标志。根据兴国县和岳西县三个时相的土壤侵蚀动态态监测图和调查统计资料分析表明，建国以来南方山地丘陵区土壤侵蚀的动态变化，基本上呈抛线形发展态势；江西、福建、湖南、广东等省的土壤侵蚀面积，均有明显下降趋势。     

付里叶变换近红外漫反射光谱法测定谷子、玉米中多种氨基酸含量

本文介绍了我们用FTNIRDRS方法对谷物样品(谷子、玉米)中多种氨基酸含量的定量分析的研究工作。预测方程的相关系数R绝大部分大于0.85,40%组分的R大于0.90。     

双通道近地遥感图像采集处理系

利用2CCD多光谱相机设计了近地遥感图像采集平台,同步获取小范围区域作物的可见光和NIR图像,通过图像处理技术从作物图像获得反射光谱信息.采集平台包括一个2CCD多光谱相机,两个采集盒,一台具有两个千兆网口的田间计算机.图像处理部分通过2GRB灰度化、中值滤波和Otsu二值化,从背景中分割出作物图像,结合作物原始R分量灰度图像,提取出作物红光图像平均灰度值,利用建立的灰度值与反射率的线性模型,计算得到作物红光波段的反射率.试验结果表明,图像采集平台工作稳定,利用图像处理方法提取的反射率与ASD光谱仪测量结果有较好相关性,为从作物冠层图像探测生长状况提供了理论依据.     

Quantifying nitrogen status of corn (Zea mays L.) in the field by reflectance measurements

Nutrient deficiencies can seriously reduce yield and economic returns to farmers. Tools that can rapidly quantify the nutritional status of plants are needed for efficient fertilizer management. Reflectance measurements have shown to be a useful tool to identify the nutritional status of different plant species. A set of calibration curves relating reflectance ratios to the nitrogen (N), phosphorus (P), magnesium (Mg), and iron (Fe) concentrations in corn leaves was established in greenhouse trials in a previous study. In this paper these calibrations were examined for their ability to identify nutrient deficiencies under field conditions. A 2-year field experiment was conducted to check and define the regions of the spectra that are influenced by leaf N concentration and to set up possible equations for quantifying the leaf N status in the field. The experiment was carried out on a loess derived soil in south-western Germany. Reflectance of corn leaves, from plants grown with six different N fertilization treatments ranging from 0 to 160 N kg ha⁻¹, was determined once a week from the beginning of June until the end of July. Reflectance measurements were performed at the 4th leaf of corn plants with a digital LEICA S1 Pro camera under controlled light conditions. Reflectance was measured in different wavelength ranges in the visible and infrared spectra. Leaf scans were evaluated within the L*a*b*-color system. Total N concentration of corn leaves was determined chemically and correlated with reflectance patterns. Significant correlations between corn N status and leaf reflectance changes were obtained at a nitrogen level of N<3.0%. Reflectance patterns at 510₇₈₀, 516₁₃₀₀, 540₁₃₀₀ nm were found most suitable to the corn N status in the field regardless of the year or sampling date. The results indicate that the spectral patterns and the defined calibration curves of N deficiency from greenhouse studies could be used in field studies. Thus, reflectance measurements may serve as a rapid, non-destructive approach to discriminate nitrogen deficiency in the field.     

利用可见光/近红外-短波红外光谱预测土壤总氮含量的研究

利用航天遥感影像,分析了影像光谱值与土壤总氮含量之间的关系.利用多元统计方法,建立了总氮含量预测模型,并制作了总氮含量分布图.