Rapid and nondestructive identification of metabolites in fresh tobacco leaves is crucial for understanding quality formation and optimizing production. This study utilized hyperspectral imaging (400-1000 nm) to analyze fresh tobacco leaves at three maturity stages (M1, M2, and M3) and identified 11 key metabolites with significant differences. Four machine learning models-Random Forest (RF), extreme gradient boosting (XGB), convolutional neural network (CNN), and long short-term memory (LSTM)-were employed to predict metabolite content. Experimental results demonstrated that CNN achieved the highest prediction accuracy, with an average R2 of 0.917, outperforming RF (0.896), XGB (0.895), and LSTM (0.915). The superior performance of CNN highlights the potential of deep learning in metabolomics research. These findings provide insights into the metabolic dynamics of tobacco maturation and offer practical applications for quality control and production optimization in the industry.

Complex environments featuring variable lighting and backgrounds similar in color to the target objects present challenges for the rapid and accurate detection of tobacco leaves, which is critical for the development of automated tobacco leaf harvesting robots. This study introduces a depth filtering approach to filter out complex regions based on distance information, thereby simplifying the detection task, and proposes a lightweight detection method based on an enhanced YOLOv5s model. Initially, the YOLOv5s backbone network is substituted with a more lightweight MobileNetV2 to reduce the model size. Subsequently, sparse model training combined with the scaling factor distribution rules of batch normalization layers is utilized to identify and eliminate inconsequential neural network channels. Finally, fine-tuning and knowledge distillation techniques are employed to achieve a model accuracy close to the YOLOv5s baseline. Experimental results indicate that the depth filtering method can improve the model's precision, recall, and mean Average Precision (mAP) by 11.2%, 29.6%, and 17.1%, respectively. The optimized lightweight model achieves a precision of 91.1%, a recall of 90.8%, and an mAP of 91.6%, with a memory footprint of only 1.4MB. It delivers a detection frame rate of 112 fps on desktop computers and 21 fps on mobile devices, which is approximately 3.5 and 4 times faster, respectively, compared to the baseline YOLOv5s tobacco leaf detection model. The precision, recall, and mAP experience a marginal decrease of 3.8, 1.6, and 2.8 percentage points, respectively, while the memory consumption is merely 10% of the pre-optimization amount. In summary, the proposed method enables the accurate detection of tobacco leaves against near-color backgrounds. Simultaneously, it achieves effective lightweighting of the model without compromising its performance, thereby providing technical support for deploying tobacco leaf detection on mobile platforms.

Accurately measuring tobacco harvest maturity is crucial for optimizing quality and crop management. Traditional methods heavily rely on qualitative evaluation and chemical experiments, which introduce subjectivity and inherent limitations. This study proposes a novel method that combines deep learning with proximal hyperspectral imaging (HSI) to achieve precise recognition of tobacco leaf maturity. Unlike traditional techniques, HSI captures rich spectral and spatial data, enabling comprehensive analysis. It adopts a pixel-level annotation strategy to annotate each pixel based on its maturity level, thereby preserving spectral complexity. The dataset comprises 3000 randomly extracted cubes (5x5x176) from 150 original hyperspectral images, encompassing five maturity levels. Spectral and spatial features are extracted from hyperspectral data using a three-dimensional convolutional neural network (3D-CNN) architecture. This method effectively leverages complex spectral patterns for maturity recognition. During testing, the model demonstrated an impressive average accuracy of 99.93%. Visual predictions vividly illustrate the model's proficiency in maturity recognition, affirming its practical utility. This study pioneers the integration of deep learning and hyperspectral near-end sensing technology in tobacco maturity assessment, mitigating the constraints of traditional methods, and establishing the groundwork for real-time monitoring and quality control of tobacco.

Microorganisms present on the surface of tobacco leaves play a significant role in shaping the composition of the tobacco microbial ecosystem, which undergoes continuous changes throughout the curing process. In the present study, a total of four distinct tobacco curing periods were selected for sampling, namely the fresh, yellowing, leaf-drying, and stem-drying stages. The bacterial 16S rRNA gene sequences of the collected samples were subsequently analyzed to identify operational taxonomic units (OTUs). The findings indicated that the complete dataset of leaf microbial samples was clustered, resulting in the identification of 1,783 operational taxonomic units (OTUs). Furthermore, the analysis of diversity revealed a pattern of initially increasing and subsequently decreasing community diversity. Redundancy Analysis (RDA) and weighted gene correlation networks for analysis (WGCNA) were employed in conjunction with environmental factors to assign OTUs to 22 modules for functional analysis. Additionally, a classification model utilizing the random forest algorithm was utilized to identify seven marker microorganisms (Escherichia coli, Faecalibacterium prausnitzii, Faecalibacterium, Escherichia-Shigella, Peptostreptococcaceae, Peptostreptococcales-Tissierellales, and Proteobacteria) that exhibited discriminative characteristics across different time periods. This study aimed to investigate the dynamic changes in the bacterial community throughout the curing process and their impact on the community’s function. Additionally, certain bacteria were identified as potential markers for detecting changes in the curing stage. These findings offer a novel opportunity to accurately regulate the curing environment, thereby enhancing the overall quality of tobacco leaf curing.

为深入解析光对烟草早期幼苗发育的调控机理,以烤烟品种CV87为试材,设置 12 h光周期和持续黑暗2个处理,采用表型组学研究了烟草种子培育第1～9天的形态变化,在此基础上采用转录组学研究了其发育第4天基因表达网络差异.结果表明,光下种子发芽势和发芽率提高约30%,子叶开口长增加约2 cm,而下胚轴长度缩短约2 cm.在早期幼苗发育阶段有551个基因差异表达受光诱导,其中364个表达上调,187个表达下调.551个基因中有28个在蛋白水平存在互作关系,其中包括AMT1-3、PSAK、LHCB5等15个上调表达基因和OLEO1、LEA7和PDF2.3等13个下调表达基因.GO富集表明,这 28 个DEGs主要与光合作用、光系统、植物激素响应有关.综上所述,本研究解析了烟草早期幼苗发育过程中响应光信号的表型变化,并通过转录组分析挖掘到一些调控表型差异相关基因,这些结果为进一步研究光环境对烟草萌发和早期幼苗发育的影响提供了参考.

对比分析了烤烟与其它主要农作物、全国不同种烟区烤烟成本收益情况,并运用DEA模型进行了生产效率评价.结果表明,2013～2018年,我国烤烟种植的亩均产值、总成本、物质与服务费用、人工成本和土地成本均呈递增趋势;土地成本涨幅最大,年涨幅3.2％;与其它作物相比,烤烟生产存在人工成本高、雇工较多、土地成本高、收益高等特点,烤烟生产的综合生产效率与其它主要农作物处于同一水平.全国20个省(直辖市)烤烟种植效率的综合效率值存在较大差异,相关省份需找准产生冗余和不足的重点,提升效率水平.提出了提高烟叶收购价格、促进土地流转、适度规模经营、推动减工降本、提升烟区的投入产出效率等对策建议.

With the increasing demand for fossil fuels, decreasing fossil fuel reserves and deteriorating global environment, humanity urgently need to explore new clean and renewable energy to replace fossil fuel resources. Biodiesel, as an environmentally friendly fuel that has attracted considerable attention because of its renewable, biodegradable, and non-toxic superiority, seems to be a solution for future fuel production. Tobacco (Nicotiana tabacum L.), an industrial crop, is traditionally used for manufacturing cigarettes. More importantly, tobacco seed is also widely being deemed as a typical inedible oilseed crop for the production of second-generation biodiesel. Advancements in raw material and enhanced production methods are currently needed for the large-scale and sustainable production of biodiesel. To this end, this study reviews various aspects of extraction and transesterification methods, genetic and agricultural modification, and properties and application of tobacco biodiesel, while discussing the key problems in tobacco biodiesel production and application. Besides, the proposals of new ways or methods for producing biodiesel from tobacco crops are presented. Based on this review, we anticipate that this can further promote the development and application of biodiesel from tobacco seed oil by increasing the availability and reducing the costs of extraction, transesterification, and purification methods, cultivating new varieties or transgenic lines with high oilseed contents, formulating scientific agricultural norms and policies, and improving the environmental properties of biodiesel.

[目的]探索不同能源类型密集烤房烘烤性能与应用效果,为烟叶绿色烘烤、节能减排和提质增效提供理论依据.[方法]以燃煤密集烤房为对照(CK),对比分析内置式生物质加热新能源密集烤房(T1)和空气源热泵+太阳能辅助加热新能源密集烤房(T2)设备成本、运行参数、故障发生次数、烤房烘烤性能、能耗用工、烘烤过程中污染物排放量,以及烤后烟叶经济性状和综合品质(颜色、柔软度、化学成分和感官质量)的差异.[结果]与CK相比,T1处理生物质烤房设备成本居中,烘烤性能较佳,T2处理烤房设备及改造成本最高,升温稳温相对较慢;同时T1和T2处理烤房设备发生故障次数相对较少.不同能源类型密集烤房能耗用工烘烤成本排序为:CK(2.63元/kg)>T1处理(2.32元/kg)>T2处理(1.19元/kg),且不同处理间差异显著(P<0.05,下同).烘烤过程中不同温度点CK总污染物(烟尘+SO2+NOX+NO)平均排放量是T1处理的4.34倍,其中在42、54和67℃时分别是T1处理的7.74、3.46和3.28倍;而T2处理烤房以电能和太阳能作为能源,烘烤过程中污染物排放量忽略不计.T1处理烤后烟叶的上等烟比例、橘黄烟比例和均价最高,杂色烟和级外烟(烘烤损失)比例最低,且T1处理上等烟比例显著高于CK和T2处理,分别提高8.01％和16.05％.CK和T1处理烤后烟叶颜色参数L值显著大于T2处理,增幅分别为7.22％和6.32％,且b值较大,即CK和T1处理烤后烟叶亮度和黄色色彩较浓;CK烤后烟叶颜色参数a值(10.64)显著小于T2处理(12.73),即T2处理烤后烟叶橘色色彩较浓.T1和T2处理烤后烟叶柔软性较好,其柔软度值显著小于CK,分别是CK的73.01％和52.08％.与CK相比,T1处理烤后烟叶化学成分协调性和感官质量评价最佳,T2处理居中.[结论]内置式生物质加热新能源密集烤房烘烤性能与应用效果较好,是当前替代燃煤密集烤房的主要烤房类型之一,在生物质燃料与相关设施配套齐全的地区适宜推广;空气源热泵+太阳能辅助加热新能源密集烤房烤后烟叶橘色色彩浓、柔软性好,生态效益好,是未来烤烟烘烤发展的重要方向.

目前烟农根据烟叶外观特征判别烟叶成熟度,具有较强的主观性强,影响烟草烘烤工艺的制定及烤后烟叶的品质.该研究以贵州省云烟87中部烟叶为研究对象,首先利用搭建的高光谱信息采集系统获取了烟叶的高光谱图像;然后应用软件SpecView提取烟叶样本感兴趣区域ROI的平均光谱,采用Savitzky-Golay平滑法预处理降噪后,将308个样本随机分208个作为训练集和100个作为测试集;最后计算了6个光谱指数及4个可见光光谱参数平均值,选择其中具有线性变化规律的7个(SIPI、mNDVI705、VOG3、I1、I2、I3、I4)光谱参数,创建了偏最小二乘法判别分析(PLS-DA)模型,模型对烟叶成熟度判别准确率达到98％.研究结果表明,将高光谱图像技术结合PLS-DA能够准确地判别烟叶的成熟度.

研究烤烟叶片叶绿素含量与高光谱参数的相关性,建立叶绿素含量估算模型,为构建或筛选系统的烟叶烘烤特性评价指标奠定基础.以云烟87为研究对象,测定不同成熟度水平和不同烘烤温度下,叶片叶绿素含量及400～1000 nm光谱反射率,以烤烟叶片高光谱反射率与烤烟叶片叶绿素含量为数据源,用SPA(连续投影算法)对高光谱数据进行特征波段筛选,筛选出10个与叶绿素含量相关的特征波长作为实验样本数据,采用基于SPA算法的SPA-BP,SPA-Ridge和SPA-LR3种预测模型预测不同烘烤温度点烟叶叶片叶绿素浓度,并比较各模型的决定系数(R2),均方根误差(RMSE)以及均方误差(MRE).3种基于SPA连续投影算法的预测模型都能较好有效预测不同烘烤温度点烟叶叶片叶绿素含量,其中SPA-BP预测模型效果最好,R2达到了0.967,RMSE为0.101,SPA-LR预测模型次之,R2达到了0.956,SPA-Ridge预测模型最低,R2达到了0.916,经验证SPA-BP预测模型的准确率为83.33％,SPA-LR预测模型的准确率为75％,SPA-Ridge预测模型的准确率为70.83％,表明BP神经网络方法的预测效果要优于线性方法,具有更好的寻优能力和预测精度,预测模型可为烟叶烘烤过程中叶绿素含量的定性研究提供理论依据.

设置3个采收成熟度(尚熟、成熟、完熟)处理,对烘烤过程烟叶叶绿素和类胡萝卜素降解规律、抗氧化酶活性和相对电导率的变化进行系统研究.结果表明,(1)各个色素组分降解主要发生在变黄中后期(24～72 h),色素降解速率和烟叶变黄速率随着成熟度的提高而提高;(2)烟叶丙二醛(MDA)含量随烘烤进程逐渐升高,在72 h时达到峰值,成熟烟叶中的MDA含量低于尚熟和完熟烟叶处理;(3)不同处理烟叶抗氧化酶[过氧化物酶(POD)、过氧化氢酶(CAT)、总超氧化物歧化酶(T-SOD)]活性均在随烘烤进程呈先升高后降低的趋势,且活性在变黄期达到峰值,成熟处理烟叶中的POD、T-SOD、CAT活性均表现出较高水平;(4)不同处理烟叶相对电导率随烘烤进程逐渐上升,完熟烟叶的相对电导率显著高于尚熟和成熟烟叶,定色中后期烟叶相对电导率明显增大,以成熟处理烟叶相对电导率最小.成熟烟叶的色素降解比例较大,MDA含量较低,抗氧化酶活性较高,相对电导率较低,说明成熟烟叶细胞膜的完整性更好,有利于烟叶细胞内部物质的生理生化转化.

为减少烤房能耗过高和污染等问题,采用新型内置一体式生物质密集烤房,与燃煤、外置式生物质密集烤房在烘烤性能、废气排放、烤后烟叶产质量、综合效益等方面进行了对比研究.结果表明,新型内置一体式生物质密集烤房稳温性能优于燃煤密集烤房,与外置式生物质烤房性能相当;综合热效率分别提高了12.54％和5.01％,干烟用工成本分别降低了88.98％和10.80％,烤后烟叶均价分别提高了1.89和0.26元/kg,烤后烟叶油分略有提升、杂气略有减轻,烟叶烘烤平均净均价分别增加2.63和0.46元/kg;中部烟叶烘烤过程中4种有害废气总含量分别降低了9235.70和1524.73 mg/m3.因此,新型内置一体式生物质密集烤房烘烤效果最佳,更具有推广价值.

烟叶含水量的快速检测在烟草种植业中起着关键的作用,检测采摘期烟叶水分含量,对烟草工艺具有重要意义.为了快速、无损地检测采摘期烟叶水分含量,提出一种主成分分析(PCA)结合马氏距离算法(MD)的方法来剔除异常样本,再使用偏最小二乘法(PLS)估测采摘期烟叶水分含量.首先,利用GaiaSky-mini2机载高光谱成像仪获取到141个采摘期烟叶的高光谱数据,采用多元散射校正(MSC)、标准正态变量交换(SNV)和Savitzky-Golay卷积平滑法等对原始光谱进行预处理.然后,应用主成分分析结合马氏距离法对校正集中的异常样品进行剔除.最后,使用偏最小二乘法(PLS)建立采摘期烟叶水分含量分析模型.结果 表明:利用SG卷积平滑法预处理的PCA-MD-PLS模型效果最佳,对烟叶含水量预测能力最好,预测模型相关系数为0.8527,均方差为1.3766.

针对烤烟油分特征预测模型的特征优选问题,提出一种改进RF(随机森林)算法特征选择策略,首先通过RF特征选择算法计算出各个特征的RF-Score,将特征按RF-Score的大小排序依次添加到特征子集中,若分类器分类准确率提高则保留该特征,若分类器分类准确率没有提高或降低则去除该特征.结果 表明:利用RF特征选择算法对烤烟高光谱特征进行筛选时,将176个高光谱特征中按基尼系数降序排列依次输入SVM分类器中,前64个高光谱波段特征即可使支持向量机分类器性能最佳,特征子集维度为64,其分类准确率为93.33％.利用改进RF特征选择策略对176个烤烟高光谱波段特征进行筛选,只需输入371.08 nm、716.71 nm、378.31 nm、487.77 nm、484.09 nm、535.85 nm六个波段的高光谱特征即可使支持向量机分类器性能最佳,其分类准确率为95％,特征子集维度为6,说明改进的RF特征选择策略可以在保证分类器性能的前提下能较好地进行数据降维,减小特征集的冗余.改进后的RF特征选择算法与全高光谱波段相比,特征数量减少170个,分类准确率提高3.33％;与RF特征选择算法相比,特征数量减少58个,分类准确率提高1.67％.

双孢蘑菇(Agaricus bisporus)又名蘑菇、白蘑菇、洋蘑菇等[1],属草腐菌,为低温型菇类,其味道鲜美,营养丰富,是世界性栽培和消费的食用菌.欧洲是栽培双孢蘑菇发源地,也是双孢蘑菇的主要栽培区域及消费市场;双孢蘑菇的栽培早已采用三次发酵技术,周年重复栽培8轮,产量稳定在30 kg/m2以上[2].而我国主要以自然季节栽培双孢蘑菇为主,受气候影响大,产量不稳定,大多每轮产量在10～15 kg/m2.

为促进烟农增收和保障优质烟叶原料供给提供科学依据,采用数据包络分析方法(DEA),对贵州烤烟与主要粮蔬作物的生产效率进行分析比较.结果 表明:烤烟、中籼稻和玉米的成本组成均以人工成本为主,分别占其总成本的59.5％、65.2％和68.2％,烤烟总成本是中籼稻和玉米的2.16倍和2.74倍;物质与服务费、人工成本和土地成本均以烤烟最高、中籼稻其次和玉米最低;烤烟的土地成本分别是中籼稻和玉米的1.99倍和3.03倍;烤烟产值和现金收益均高于中籼稻和玉米,净利润、现金收益低于马铃薯、菜椒、大白菜和西红柿4种蔬菜作物;贵州烤烟的综合生产效率小于1,与中籼稻和玉米为同一水平,在降低一定生产成本投入和增加现金收益886.71元/667m2的情况下,可达到DEA有效水平.结合贵州省烤烟产业的实际情况,提出了提升其生产效率的建议.

为了在提高烟叶烟碱含量的同时控制总氮、蛋白质含量,通过在现蕾前足叶打顶并采用低施氮(45 kg/hm2、67.5 kg/hm2和90 kg/hm2)、少留叶(12片/株、14片/株、16片/株)的措施来探索其对K326烟叶烟碱、总氮等化学成分的影响.结果表明:(1)所有低施氮少留叶处理的烟叶其烟碱含量和烟碱氮/总氮均有所提高,同时蛋白质含量下降,总氮和糖含量较为适宜.(2)施氮量对烟叶的烟碱、总氮、蛋白质、氯含量有极显著影响,留叶数对烟叶的烟碱、总糖、总氮含量和烟碱氮/总氮有极显著影响,施氮量×留叶数交互作用对各个部位化学成分的影响均不显著.(3)施氮量和留叶数对不同部位烟叶化学成分的影响程度不同.采用低施氮少留叶措施可提高烟叶烟碱含量和烟碱氮/总氮,降低蛋白质含量;施氮量主要影响了烟叶的烟碱、总氮、蛋白质和氯含量,留叶数主要影响了烟叶的烟碱、总氮、总糖含量和烟碱氮/总氮,二者的交互作用对化学成分影响不显著.

为探明烤烟种植中适宜的施氮量及留叶数,采用双因素裂区试验研究低施氮量(氮用量为45 kg/hm2、67.5 kg/hm2和90 kg/hm2,常规施氮量为105 kg/hm2)和少留叶数(烟叶留叶数为12片/株、14片/株、16片/株,常规留叶数为18片/株)对不同部位烟叶烟碱氮、蛋白氮、总氮含量及感官质量的影响.结果 表明:1)在低施氮及少留叶条件下,施氮量对烟叶烟碱氮、蛋白氮和总氮含量影响显著,对香气质、香气量、吃味、杂气、刺激性及总分影响极显著;留叶数对烟叶烟碱氮影响极显著,对总氮含量、香气质和香气量影响显著;施氮量与留叶数的交互作用对含氮化合物和感官质量影响不显著;2)与常规栽培(施氮量105 kg/hm2+留叶数18片/株)相比,所有低施氮、少留叶处理初烤烟叶的蛋白氮含量均有所下降,烟碱氮含量(施氮量45 kg/hm2+留叶数16片/株的处理组合除外)均有所提高;3)烟碱氮/总氮比值在0.305～0.315,香气质、香气量和吃味得分均随着烟碱氮/总氮比值的增加而增加.施氮量67.5 kg/hm2+留叶数16片/株处理组合的化学成分较协调,感官质量最好;4)低施氮、少留叶处理的总糖和还原糖含量较为适宜,可提高烟叶烟碱氮含量,降低蛋白氮含量;施氮量对烟叶含氮化合物和感官质量影响最大,留叶数次之.

The 3D interactive imitation software for tobacco curing training was simulated by using computer technology to reduce the tobacco curing training cost,break the time and space limit of tobacco curing training and shorten training time.The software with a new interactive media training pattern can replace the conventional planar curing training pattern,which breaks the time and space limit,effectively enhances working efficiency and realizes the training objective of tobacco farmers and curing technicians without delaying tobacco curing.The software fills the gap in software application in tobacco curing training at the same time.

To enhance the usability of bottom leaves with low commercial value,the chemical components and smoking quality of bottom leaves with different harvesting time and leaf numbers were analyzed.Results:Picking 4 leaves in 80th d (T5),picking 4 leaves in 80th d (T6),picking 4 leaves in 90th d (T8) and picking 6 leaves in 90thd (T9) were of good quality,sufficient aroma quantity,proper concentration,little irritation and comfortable aftertaste;The chemical composition of which were of higher nicotine,higher sugar,lower nitrogen and lower chlorine.Harvesting time significantly influenced the chemical content in the usability of bottom leaves.The preliminary analysis indicated that the quality of harvesting 4 ～ 6 leaves during the field period 80th d,90thd (harvesting leaves) was beneficial for enhancing the usability of bottom leaves with low commercial value.