湖北三峡库区晚熟脐橙不同海拔果实品质变化及枯水防控技术研究

以湖北三峡库区秭归地区不同海拔伦晚脐橙（Citrus sinensis Osbeck ‘Lane late’ Navel）果实为试验材料，探讨了湖北三峡库区晚熟脐橙果实枯水防控栽培技术，进行了晚熟脐橙种植区域布局研究，结果表明，在湖北三峡库区，江南晚熟脐橙适宜区域化种植在海拔360 m以下，江北在海拔350 m以下。在适宜范围外，随海拔升高，晚熟脐橙果实品质下降，可食率、果汁率降低，酸下降快，化渣性变差，果实枯水加重。其次，探讨了晚熟脐橙果实枯水防控生产技术，冬季套袋试验效果好，果实枯水比率控制在5%以下，显著低于对照果实（枯水率50%），可有效减轻和防止晚熟脐橙果实枯水，而且套袋果实外观好，品质优。覆膜技术效果其次，果实枯水比率可以控制在20%以下，而且该技术简便、劳动量小且成本较低，可以在生产中应用。     

基于迁移学习的GoogLeNet果园病虫害识别与分级

为提高果树病虫害危害程度分级精度进而更好地指导果园病虫害防治，采用迁移学习技术与GoogLeNet模型相结合的方法，对6种果园作物的25类病虫害样本进行识别与危害程度分级研究；同时，探究不同数据集大小以及不同优化算法对模型性能的影响；基于MATLAB平台设计了一款可视化的病虫害识别与分级系统。结果表明:1)基于迁移学习的GoogLeNet模型，对病虫害识别精度可达99.35%，危害程度分级精度可达92.78%；2)在相同训练参数下，本研究模型比AlexNet、VGG-16、ResNet-18、SqueezeNet、原GoogLeNet及MobileNet-v2模型验证精度提高了2.38%~11.44%，并且收敛速度最快；3)本研究模型识别精度随着数据集的增大而提高；在3种优化算法中SGDM算法耗时最短且精度最高，更适合本研究模型。通过拍摄果树叶片病害区域图像，本研究设计的系统能够在0.43 s左右准确识别出果树种类、病害类型以及危害等级等信息。     

落叶果树冬季管理技术要点

落叶果树冬季进入休眠期，果农田间管理时间长，农事操作方便，而在生产中，不少果农忽视了冬季这个果园管理的关键时期，造成树势衰弱、树形杂乱、病虫害严重。做好落叶果树冬季管理技术措施，对于增强树势，提高翌年果树产量、品质和效益起到关键性的作用。     

基于逆运动学降维求解与YOLO v4的果实采摘系统研究

为提高采摘设备的执行效率，采用六自由度机械臂、树莓派、Android手机端和服务器设计了一种智能果实采摘系统，该系统可自动识别不同种类的水果，并实现自动采摘，可通过手机端远程控制采摘设备的起始和停止，并远程查看实时采摘视频。提出通过降低自由度和使用二维坐标系来实现三维坐标系中机械臂逆运动学的求解过程，从而避免了大量的矩阵运算，使机械臂逆运动学求解过程更加简捷。利用Matlab中的Robotic Toolbox进行机械臂三维建模仿真，验证了降维求解的可行性。在果实采摘流程中，为了使机械臂运动轨迹更加稳定与协调，采用五项式插值法对机械臂进行运动轨迹规划控制。基于Darknet深度学习框架的YOLO v4目标检测识别算法进行果实目标检测和像素定位，在Ubuntu 19.10操作系统中使用2000幅图像作为训练集，分别对不同种类的果实进行识别模型训练，在GPU环境下进行测试，结果表明，每种果实识别的准确率均在94%以上，单次果实采摘的时间约为17s。经过实际测试，该系统具有良好的稳定性、实时性以及对果实采摘的准确性。     

基于多分类器DS证据理论融合的水果识别研究

针对不同分类器对不同水果种类识别准确率的不均衡问题,提出一种基于多分类器DS证据理论融合的水果识别方法。该研究选择kaggle上fruits360数据集中的5种水果作为研究对象,首先对预处理后的5种水果图像的颜色、纹理、形状特征进行提取,分别选用BP神经网络、K均值、SVM三种分类器,结合被测图像在每种分类器上的识别结果和各个分类器对不同水果的分类准确率,构建基本概率函数(BPA函数),通过DS证据融合规则对分类器融合后对被测图像进行识别。试验结果表明:该方法对5种水果的识别平均准确率为95.2%,总体标准偏差为0.02993,在提高单分类器识别准确率的同时,解决了分类器对各种水果识别的不均衡问题。对10组测试集识别的平均准确率为93.5%,总体标准偏差为0.055,该方法对水果种类的识别更准确和稳定。     

安康市核桃果实害虫种类调查研究

采用生长期定点定株定期果实害虫调查和贮藏期室内常温贮藏观测法,对安康市核桃果实害虫种类进行了调查。结果为安康市核桃果实生长期害虫有13种,隶属5目11科;核桃果实贮藏期害虫有3种,隶属2目3科;核桃生长期危害较大的果实害虫主要是核桃举肢蛾、核桃长足象、双鬃尖尾蝇和桃蛀螟;贮藏期危害较大的果实害虫是印度谷螟;首次发现并记录了角斑古毒蛾幼虫为害核桃幼果。     

Factor Analysis and Comprehensive Evaluation of the Fruit Quality of ‘Yangjiaocui’ Melons

【Objective】The aim of this study was to identify the typical indices of fruit quality and high quality materials of ‘Yangjiaocui’ melons to provide reference and basis for melon quality evaluation and breeding. 【Method】Eighteen indices, including economic trait and flesh quality of fruits from 6 cultivars and 26 high-generation inbred lines of ‘Yangjiaocui’ melons were determined. Based on factor analysis, stepwise regression, correlation analysis and two-dimension ordination map, the typical indices of fruit quality and high quality materials were identified. 【Result】 Among the 18 indices, single fruit mass, hardness, vitamin C (V_C) content, titratable acid content, sugar-acid ratio and solid acid ratio showed greater variability, and the variation coefficients were more than 20%. However, the variability of fruit transverse diameter, flesh thickness rate and pH were relatively lower and the variation coefficients were less than 10%. Three common factors were extracted from economic trait indices with 87.71% cumulative variance contribution rate and the explanatory indices were fruit longitudinal diameter, fruit transverse diameter, fruit shape index, fruit cavity longitudinal diameter, fruit cavity transverse diameter and flesh thickness rate. From quality indices, 4 common factors were extracted with 90.03% cumulative variance contribution rate and the explanatory indices were V_C content, hardness, total soluble solids content, total sugar content, titratable acid content, sugar-acid ratio and solid-acid ratio. Based on the comprehensive analysis of these selected 6 economic trait indices and 7 flesh quality indices, 5 common factors were extracted with 87.35% cumulative variance contribution rate. According to the correlation among the indices and the principle of simplicity and ease of use and avoidance of overlapping information, 5 indices, including fruit longitudinal diameter, flesh thickness rate, hardness, total sugar content and titratable acid content, could be used as typical indices to evaluate the fruit quality of ‘Yangjiaocui’ melons. The comprehensive quality score of all materials were ranked depended on scores and variance contribution rate of the above 5 common factor and two-dimension ordination map. Finally, 10 materials were selected as high quality materials. 【Conclusion】 Fruit longitudinal diameter, flesh thickness rate, hardness, total sugar content and titratable acid content could be used as typical indices to evaluate fruit quality of ‘Yangjiaocui’ melons. Inbred lines 25, 24, 22, 21, 8, 1, 17, 18, 20 and 23 were high quality materials.     

Molecular identification of blast resistance genes in rice landraces from northeastern India

Rice blast disease caused by the fungus Magnaporthe oryzae is one of the most devastating diseases causing huge losses worldwide. In the present study, major blast resistance genes were investigated in landraces originating from northeastern India. Based on phenotypic evaluation, 288 landraces were classified into three distinct groups: resistant (75), moderately resistant (127) and susceptible (86). The genetic frequencies of the 18 major blast resistance genes were between 6.2% and 27.4%, with only two genotypes possessing a maximum of nine blast resistance genes. The cluster and population structure analysis grouped the landraces into two groups. Through principal coordinate analysis, the scatter plots partitioned the resistant and moderately resistant landraces into different groups. Analysis of molecular variance showed maximum (96%) diversity within populations and least (4%) diversity between populations. Association analysis identified six markers, CRG4_2, RM72, tk59-2, pi21_79-3, RM1233 and RM6648, that are significantly associated with blast disease and explained a phenotypic variance of 1.1–6.5%. The associated genes could be used in marker-assisted rice breeding programmes for gene pyramiding to develop rice varietal resistance against blast disease. The present study represents a valuable blast resistance genetic resource that could be used for identification of new R genes, donors for blast resistance, and genomic studies.     

The Identification of QTL Associated with Cotton Fruit Branch Length Suitable for Mechanized Harvest Utilizing BSA-seq

[Objective] This study aims to identify quantitative trait loci (QTLs) associated with cotton fruit branch length by using the combination of bulked-segregant analysis (BSA) and next-generation sequencing (NGS) method. [Method] In this study, the cotton F₂ segregating population was developed with the short branch line Sujimian 125 and long-branch variety Sikang 1 as cross parents. According to the average internode length of fruit branches in middle, extreme individuals in F₂ population were divided into two groups to generate the bulked DNA samples. The whole-genome resequencing of two DNA bulks was applied to analyze the single nucleotide polymorphism (SNP) and insertion-deletion (InDel) between two groups, and the fruit branch internode length related QTLs were detected using Euclidean distance (ED). [Result] With SNP data obtained from sequencing, a 0.8 Mbp range associated region on chromosome A3 was identified; with InDel data obtained from sequencing, a 1.09 Mbp range associated region on chromosome A3 was identified. Two regions overlapped, and the overlapped range was 0.77 Mbp. [Conclusion] This result suggested that the difference between I-type fruit branch and zero-type was due to different genetic locus and pattern rather than dosage effect of the same gene.     

基于经济效益与生态效益的食用菌产业优化评价

为了优化食用菌产业结构,促进食用菌产业的可持续发展。给出了调整优化食用菌产业结构、促进食用菌产业生态经济建设的一些措施;提出重视经济效益和生态效益,合理引导发展中小食用菌企业的思路。对食用菌产业优化和生态经济的发展具有一定的参考价值。