本文主要研究内容
作者徐夺花（2019）在《基于高光谱成像技术的典型水果冲击损伤的研究》一文中研究指出：水果富含维生素、矿物质和膳食纤维,历来受到人们的青睐,并且已经成为人类饮食结构的重要组成部分。但是,新鲜水果在收获、包装和运输期间非常容易受到机械损伤,这可能导致果实品质的显著降低。在各种形式的损伤中,冲击损伤最严重且最容易发生。如果发生损伤,基于损伤的程度,通过把水果分级处理可以减少经济损失。这时就需要对损伤的程度进行客观和定量的评估。然而,这在食品安全领域仍是一个重要挑战。传统的人工感官检测和破坏性预测评价方法费时费力,而且精度不高,已经无法满足现代化农业快速、无损、实时的自动检测分级的需求。本文分别以芒果和苹果等典型水果为研究对象,基于高光谱成像技术,并结合多种光谱分析技术及数学建模方法,研究了水果受冲击损伤后其光谱的变化、理化指标的变化和相应力学参数的变化,寻找影响损伤程度的主要贡献因子,进一步建立了果实光谱信息与其理化指标、力学参数的数学模型,实现了果实是否受损的无损鉴别及对其品质等各参数的预测,为无损评估果实的机械损伤提供重要参考。主要研究内容和成果有:(1)采用近红外高光谱成像技术(900-1700nm)实现了对芒果理化指标的预测及损伤程度的无损鉴别。采集从不同高度(0.5m,1.0m,1.5m)跌落的芒果样品的高光谱图像,提取样品感兴趣区域(Region of interest,ROI)的平均光谱。结合竞争性自适应重加权算法(Competitive adaptive reweighted sampling,CARS)提取总光谱的特征波长作为光谱变量。采用理化方法测定所有样品的理化指标值。建立样品光谱变量与其相应的理化指标的偏最小二乘回归(Partial least square regression,PLSR)模型,针对果肉硬度(Pulp firmness,PF)、可溶性固形物(Total soluble solids,TSS)、可滴定酸(Titratable acidity,TA)及颜色(?b*),PLSR预测模型的决定系数~2和均方根误差RMSEP分别可达到0.84和3.16 N,0.9和0.49 ~oBrix,0.86和0.07%,0.94和0.96。最后,根据成熟度指标RPI和判别分析(Discriminant analysis,DA)对样品受冲击损伤的程度进行分类且分类结果的准确率不低于77.8%。(2)采用近红外高光谱成像技术(900-1700nm)实现了对苹果冲击损伤面积和冲击过程中力学参数的量化及预测。采集从不同高度(0.5m,1.0m,1.5m)跌落的苹果样品的高光谱图像,提取样品感兴趣区域的平均光谱。结合回归系数(Regression coefficients,RC)方法提取总光谱的特征波长。借助感压胶片技术和高速相机获得苹果跌落过程中的损伤面积、接触载荷和吸收能量。从苹果的穿刺试验提取样品的果肉硬度。统计分析表明,样品的吸收能量、接触载荷分别和其损伤面积之间呈显著线性相关,其决定系数~2分别为0.93和0.92,表明力学参数可以很好地表征苹果的冲击损伤。最后,分别建立光谱变量与其损伤面积及力学参数的偏最小二乘回归模型。针对损伤面积、吸收能量、接触载荷及果肉硬度,PLSR预测模型的决定系数~2均方根误差RMSEP可分别达到0.8和116.73mm~2,0.89和0.075J,0.53和67.38N,0.65和19.99g。
Abstract
shui guo fu han wei sheng su 、kuang wu zhi he shan shi qian wei ,li lai shou dao ren men de qing lai ,bing ju yi jing cheng wei ren lei yin shi jie gou de chong yao zu cheng bu fen 。dan shi ,xin xian shui guo zai shou huo 、bao zhuang he yun shu ji jian fei chang rong yi shou dao ji xie sun shang ,zhe ke neng dao zhi guo shi pin zhi de xian zhe jiang di 。zai ge chong xing shi de sun shang zhong ,chong ji sun shang zui yan chong ju zui rong yi fa sheng 。ru guo fa sheng sun shang ,ji yu sun shang de cheng du ,tong guo ba shui guo fen ji chu li ke yi jian shao jing ji sun shi 。zhe shi jiu xu yao dui sun shang de cheng du jin hang ke guan he ding liang de ping gu 。ran er ,zhe zai shi pin an quan ling yu reng shi yi ge chong yao tiao zhan 。chuan tong de ren gong gan guan jian ce he po huai xing yu ce ping jia fang fa fei shi fei li ,er ju jing du bu gao ,yi jing mo fa man zu xian dai hua nong ye kuai su 、mo sun 、shi shi de zi dong jian ce fen ji de xu qiu 。ben wen fen bie yi mang guo he ping guo deng dian xing shui guo wei yan jiu dui xiang ,ji yu gao guang pu cheng xiang ji shu ,bing jie ge duo chong guang pu fen xi ji shu ji shu xue jian mo fang fa ,yan jiu le shui guo shou chong ji sun shang hou ji guang pu de bian hua 、li hua zhi biao de bian hua he xiang ying li xue can shu de bian hua ,xun zhao ying xiang sun shang cheng du de zhu yao gong suo yin zi ,jin yi bu jian li le guo shi guang pu xin xi yu ji li hua zhi biao 、li xue can shu de shu xue mo xing ,shi xian le guo shi shi fou shou sun de mo sun jian bie ji dui ji pin zhi deng ge can shu de yu ce ,wei mo sun ping gu guo shi de ji xie sun shang di gong chong yao can kao 。zhu yao yan jiu nei rong he cheng guo you :(1)cai yong jin gong wai gao guang pu cheng xiang ji shu (900-1700nm)shi xian le dui mang guo li hua zhi biao de yu ce ji sun shang cheng du de mo sun jian bie 。cai ji cong bu tong gao du (0.5m,1.0m,1.5m)die la de mang guo yang pin de gao guang pu tu xiang ,di qu yang pin gan xing qu ou yu (Region of interest,ROI)de ping jun guang pu 。jie ge jing zheng xing zi kuo ying chong jia quan suan fa (Competitive adaptive reweighted sampling,CARS)di qu zong guang pu de te zheng bo chang zuo wei guang pu bian liang 。cai yong li hua fang fa ce ding suo you yang pin de li hua zhi biao zhi 。jian li yang pin guang pu bian liang yu ji xiang ying de li hua zhi biao de pian zui xiao er cheng hui gui (Partial least square regression,PLSR)mo xing ,zhen dui guo rou ying du (Pulp firmness,PF)、ke rong xing gu xing wu (Total soluble solids,TSS)、ke di ding suan (Titratable acidity,TA)ji yan se (?b*),PLSRyu ce mo xing de jue ding ji shu ~2he jun fang gen wu cha RMSEPfen bie ke da dao 0.84he 3.16 N,0.9he 0.49 ~oBrix,0.86he 0.07%,0.94he 0.96。zui hou ,gen ju cheng shou du zhi biao RPIhe pan bie fen xi (Discriminant analysis,DA)dui yang pin shou chong ji sun shang de cheng du jin hang fen lei ju fen lei jie guo de zhun que lv bu di yu 77.8%。(2)cai yong jin gong wai gao guang pu cheng xiang ji shu (900-1700nm)shi xian le dui ping guo chong ji sun shang mian ji he chong ji guo cheng zhong li xue can shu de liang hua ji yu ce 。cai ji cong bu tong gao du (0.5m,1.0m,1.5m)die la de ping guo yang pin de gao guang pu tu xiang ,di qu yang pin gan xing qu ou yu de ping jun guang pu 。jie ge hui gui ji shu (Regression coefficients,RC)fang fa di qu zong guang pu de te zheng bo chang 。jie zhu gan ya jiao pian ji shu he gao su xiang ji huo de ping guo die la guo cheng zhong de sun shang mian ji 、jie chu zai he he xi shou neng liang 。cong ping guo de chuan ci shi yan di qu yang pin de guo rou ying du 。tong ji fen xi biao ming ,yang pin de xi shou neng liang 、jie chu zai he fen bie he ji sun shang mian ji zhi jian cheng xian zhe xian xing xiang guan ,ji jue ding ji shu ~2fen bie wei 0.93he 0.92,biao ming li xue can shu ke yi hen hao de biao zheng ping guo de chong ji sun shang 。zui hou ,fen bie jian li guang pu bian liang yu ji sun shang mian ji ji li xue can shu de pian zui xiao er cheng hui gui mo xing 。zhen dui sun shang mian ji 、xi shou neng liang 、jie chu zai he ji guo rou ying du ,PLSRyu ce mo xing de jue ding ji shu ~2jun fang gen wu cha RMSEPke fen bie da dao 0.8he 116.73mm~2,0.89he 0.075J,0.53he 67.38N,0.65he 19.99g。
论文参考文献
论文详细介绍
论文作者分别是来自天津商业大学的徐夺花，发表于刊物天津商业大学2019-07-05论文，是一篇关于高光谱成像论文,水果论文,冲击损伤论文,理化指标论文,力学参数论文,偏最小二乘回归论文,竞争性自适应重加权算法论文,天津商业大学2019-07-05论文的文章。本文可供学术参考使用，各位学者可以免费参考阅读下载，文章观点不代表本站观点，资料来自天津商业大学2019-07-05论文网站，若本站收录的文献无意侵犯了您的著作版权，请联系我们删除。
标签：高光谱成像论文; 水果论文; 冲击损伤论文; 理化指标论文; 力学参数论文; 偏最小二乘回归论文; 竞争性自适应重加权算法论文; 天津商业大学2019-07-05论文;