棉花中白色异性纤维的激光成像快速检测方法

现有的棉花异性纤维分拣机采用可见光照明成像,难以识别棉花中混杂的白色异性纤维。为了有效地检测出与棉花颜色相同或相近的白色异性纤维,该文采用线激光照明成像的方法,在固定线激光功率、波长和相机光圈的条件下,利用不同的曝光时间,获取了12种典型的白色异性纤维与皮棉的图像,分析其成像曝光时间与图像对比度之间的关系,发现不同的关系曲线存在一段共有的最优曝光峰值时间。在此曝光时间内,同一图像中的白色异性纤维已经"过曝",而棉花还处于欠饱和状态。二者图像灰度值的明显差异可用于检测棉花中的白色异性纤维。该文在线激光功率为0.8 W,波长650 nm,相机光圈为8C,曝光时间为1.6 ms的条件下获取了1 500幅图像。试验表明,采用简单的固定阈值法,皮棉中12种典型的白色异性纤维的识别率达到了95.8%。研究结果为提高棉花中白色异性纤维的识别率和速度提供了一条新途径。     

高光谱图像技术在掺假大米检测中的应用

为了有效判别出优质大米中是否掺入劣质大米,该文研究了一种针对大米掺假问题的快速、无损检测方法。从市场上购买了东北长粒香大米和江苏溧水大米,按纯东北长粒香大米、3∶1、2∶2、1∶3和纯江苏溧水大米共5个掺合水平进行大米试验样本的制备。利用可见-近红外高光谱图像采集系统(390~1050 nm)获取了200个大米样本的高光谱图像。采用ENVI软件确定高光谱图像的感兴趣区域(region of interest,ROI),并提取出所有样本在ROI内的平均高光谱数据。采用支持向量机(support vector machine,SVM)建立全光谱波段下的大米掺假判别模型,径向基(radial basis function,RBF)核函数模型交叉验证准确率为93%、预测集正确率为98%。由于高光谱信息量大、冗余性强且受噪声的影响较大,该文采用主成分分析方法(principal component analysis,PCA)分别对大米高光谱图像和高光谱数据进行处理,从特征选择和特征提取2个角度对原始高光谱数据进行处理,通过主成分权重系数图选择了531.1、702.7、714.3、724.7、888.2和930.6 nm 6个特征波长,通过留一交叉验证法(leave-one-out cross-validation,LOOCV)确定并提取出PCA降维后的最优主成分数(number of principal component,PCs)为9。最后分别将优选出的特征波长和提取出的最优主成分数作为模型的输入,建立SVM模型。试验结果表明,基于特征波长SVM模型的交叉验证准确率为95%、预测集正确率为96%,基于最优主成分数SVM模型的交叉验证准确率为94%、预测集正确率为98%。该研究结果表明,该文建立的基于特征波长和基于最优主成分数的SVM模型均具有较优的预测性能,且利用高光谱图像技术对大米掺假问题进行检测是可行的。     

应用可见/近红外高光谱成像测定鲑鱼片脂肪含量分布(英文)

脂肪作为一种重要的品质参数,在大西洋鲑鱼片中的分布很不均匀。为寻找一种能替代脂肪化学检测的快速无损的方法,该研究应用可见/近红外高光谱成像测定大西洋鲑鱼片的脂肪含量分布。分别采用可见/短波近红外(400-1100 nm)和近红外(900-1700 nm)系统获取大西洋鲑鱼片样本的高光谱图像。提取样本图像的平均光谱并与其相应的脂肪含量化学值采用偏最小二乘回归(partial least squares regression,PLSR)和最小二乘支持向量机(least-squares support vector machines,LS-SVM)建立相关性模型。为降低高光谱图像的共线性和冗余度,基于竞争性自适应重加权算法(competitive adaptive reweighted sampling,CARS)分别在可见/短波近红外和近红外光谱区间提取16个(468,479,728,734,785,822,863,890,895,899,920,978,1005,1033,1040,1051 nm)和15个(975,995,1023,1047,1095,1124,1167,1210,1273,1316,1354,1368,1575,1632,1661 nm)特征波长,并分别建立PLSR和LS-SVM模型。特征波长模型的性能优于全波段模型,且近红外区间的特征波长PLSR模型为最优,预测决定系数(R2p)为0.92,预测均方根误差(root mean square error of prediction,RMSEP)为0.92%,剩余预测偏差(residual predictive deviation,RPD)为3.50。最后,将最优模型用于预测高光谱图像上所有像素点的脂肪含量以展示样本上脂肪的分布。此外,还基于该技术对大西洋鲑整鱼片实现了脂肪分布可视化。结果表明高光谱成像技术结合化学计量学方法在大西洋鲑鱼片脂肪的定量和分布可视化上有一定的研究和应用前景。     

基于HS-SPME-GC-MS的青梅酒香气成分研究

[目的]为更好地勾兑青梅酒,研究青梅浸泡酒、青梅汁、青梅汁发酵酒以及浸泡基酒的香气成分.[方法]采用顶空固相微萃取(HS-SPME)与气相色谱质谱(GC-MS)相结合的方法定性分析浸泡基酒、青梅浸泡酒、青梅汁及青梅汁发酵酒4种样品的香气成分.[结果]试验显示,4个样品依次检测出14、32、17、46种香气成分,香气成分本质的差异决定了其特有的风味和口感.其中,青梅在浸泡过程中产生的特有香气成分为苯甲醛、S-(-)-2-甲基-1-丁醇、苯甲酸乙酯、5-羟甲基糠醛等.青梅汁中特有的香气成分为糠醛、苯乙醇、苯甲醇等.青梅汁发酵酒中产生的主要微量香气成分有苯乙醇(6.941％,峰面积相对百分含量,下同)、异戊醇(6.940％)、辛酸乙酯(3.734％)、癸酸乙酯(2.590％)、己酸乙酯(2.479％)、9-癸烯酸乙酯(2.080％)、5-羟甲基糠醛(1.756％)等.[结论]研究可为青梅酒的品质改善提供科学依据和数据参考.     

Water penetration into rice grains during soaking observed by gradient echo magnetic resonance imaging

The water penetration into a rice grain during soaking was monitored by three-dimensional gradient echo magnetic resonance imaging, which has the advantage of a short measurement time and a sufficient sensitivity for low-moisture samples. The water penetration route was compared between the milled and brown rice grains of two japonica cultivars. In normal milled rice grains of Koshihikari, water first penetrated the ventral side surface and the embryo attachment site of the endosperm, then migrated along the central line and transverse cracks, and finally diffused to all parts of the endosperm. In milled rice grains of Yamadanishiki with a white core, water quickly infiltrated into the cracks or chalky parts on the dorsiventral line and then diffused to the lateral side. In brown rice grains of both cultivars water penetrated extremely slowly and did not infiltrate directly into transverse cracks or the white core due to inhibition by the pericarp and seed coat. The central part of every grain allowed more water to penetrate. The route, pattern and speed of water penetration were determined by the morphological structure, crack formation and hardness distribution associated with the packing of the starch granules in the grain.     

利用近红外高光谱图像技术快速鉴别西瓜种子品种

为了研究采用近红外高光谱图像技术对西瓜种子品种快速无损鉴别的可行性,该文采用近红外高光谱图像技术,通过提取西瓜种子的光谱反射率,结合Savitzky-Golay (SG)平滑算法,经验模态分解算法（empirical mode decomposition,EMD）和小波分析（wavelet transform,WT）对提取出的光谱数据进行去除噪声处理,采用连续投影算法（successive projections algorithm,SPA）和遗传-偏最小二乘法（genetic algorithm-partial least squares, GA-PLS）进行特征波长选择。基于全波段光谱建立了偏最小二乘判别分析（partial least squares-discriminant analysis,PLS-DA）,基于特征波长建立了反向传播神经网络（back-propagation neural network,BP NN)判别模型和极限学习机（extreme learning machine,ELM）判别模型。试验结果表明,基于特征波长的BPNN模型和ELM模型的结果优于基于全部波长的PLS-DA模型,基于SG预处理光谱提取的特征波长建立的ELM模型取得最优的判别效果,建模集和预测集的判别正确率均为100%。结果表明应用近红外高光谱成像技术对西瓜种子品种鉴别是可行的,为西瓜种子的品种快速鉴别提供了一种新方法。     

基于改进湍流模型和偏振成像技术的水下退化图像复原方法

为了满足工厂化水产养殖中过程控制的信息化要求,该文提出一种基于改进湍流模型结合偏振成像技术,具有较强鲁棒性的水下退化图像复原方法。考虑内外尺度对波结构函数影响,结合折射率谱,改进水下湍流退化模型以提高复原算法先验知识的完备性;基于改进的退化模型和水下前向散射光的偏振特性,利用偏振成像技术提取退化图像中的噪声特征;基于退化图像噪声特征,采用约束最小二乘滤波法进行退化图像复原;最后,对复原效果进行相应比较,结果表明在强湍流条件下本算法具有更为理想的复原效果。该研究可为复杂水流条件下水下退化图像复原方法研究提供参考。     

EX VIVO MAGNETIC RESONANCE IMAGING OF THE DISTAL ROW OF EQUINE CARPAL BONES: ASSESSMENT OF BONE SCLEROSIS AND CARTILAGE DAMAGE

The distal row of carpal bones (C2, C3, and C4) from eight left intercarpal joints--four from Standardbred Trotters and four from Swedish Warmblood horses--were used to assess the potential of magnetic resonance (MR) imaging to detect cartilage and bone lesions. The joints used in the study were classified by macroscopic and radiographic examinations as having normal, mild, moderate, or severe articular cartilage lesions and bone sclerosis. Those classifications correlated well with the appearance of the MR images. Bone sclerosis in the MR images was observed as regions of decreased signal intensity. Upon quantitative analysis of the MR images there was a significant difference (p < 0.0001) in the MR signal intensity from areas where radiographic bone sclerosis was observed compared to areas of radiographic nonsclerotic bone. In addition, the MR images were used to pilot the location of histology slices through areas of interest that were then examined microscopically; hence, the lesions found from the MR imaging examination were verified microscopically. It was concluded that cartilage lesions and cartilage loss are related to the sclerotic state of the underlying bone. The MR protocols developed in this study were applied on five intact cadaveric carpal joints, and it was concluded that MR imaging could successfully be used in the intact joint to detect minor cartilage and bone lesions not visualized by either radiography or macroscopic examination. Hence, MR imaging can be used to delineate interactions between articular cartilage and subchondral bone over time and in vivo.     

EVALUATION OF THE VARIABLY OSSIFIED COLLATERAL CARTILAGES OF THE DISTAL PHALANX AND ADJACENT ANATOMIC STRUCTURES IN THE FINNHORSE WITH COMPUTED TOMOGRAPHY AND MAGNETIC RESONANCE IMAGING

Six Finnhorse cadaver forefeet were selected to represent radiographically different types and grades of ossification of the collateral cartilages of the distal phalanx. These cartilages and adjacent tissues were evaluated with computed tomography (CT) and high field magnetic resonance imaging (MRI). In CT the internal structure of the cartilages was consistent, but in MRI some differences were noted. The shape of the collateral cartilages and their ligamentous attachments varied. The border between ossified and non-ossified cartilage appeared distinct, with considerable variation in the extent of the ossified area in regard to the cross-sectional area of the cartilage. Ossification originating from the palmar processes and extending in the proximaVpalmaroproximal direction, without separate centers of ossification, generally appeared smooth and inactive. Palmar ossification followed the irregular shape of the cartilage. Separate centers of ossification had a medullary cavity or were sclerotic. Presence ofamedullary cavity or sclerosis were also found at the base of the cartilages. The incomplete fusion lines between separate centres of ossification and the ossified base of the cartilage varied from congruent and inactive to reactive with marked sclerosis, flared margins and parachondral changes. Incomplete fusion may be clinically significant. Local conformational adaptations of the hoof were also documented with extensive ossification of the collateral cartilage.     

A COMPARISON OF RADIOGRAPHY, COMPUTED TOMOGRAPHY, AND MAGNETIC RESONANCE IMAGING FOR THE DIAGNOSIS OF PALMAR PROCESS FRACTURES IN FOALS

The relative sensitivity of radiography, computed tomography, and magnetic resonance imaging for detecting palmar process fractures of the distal phalanx in foals was determined and the imaging findings were compared with histomorphologic evaluations of the palmar processes. Compared to radiography, computed tomography and magnetic resonance imaging did not improve the sensitivity for detection of palmar process fractures. Statistical agreement for palmar process fracture diagnosis was excellent among the three imaging modalities. Histomorphologic evaluations were more sensitive for diagnosis of palmar process fracture than any of the imaging modalities. Three-dimensional image reconstructions and volume measurements of distal phalanges and palmar process fracture fragments from computed tomography studies provided more complete anatomical information than radiography. Magnetic resonance imaging confirmed that the deep digital flexor tendon insertion on the distal phalanx is immediately axial to the site where palmar process fractures occur, and differentiated cartilage, bone, and soft tissue structures of the hoof.