马铃薯薯饼原料搅拌机的设计

在马铃薯薯条生产过程中会产生大量边角料和断条,这些废料可以加以利用,加工成薯饼、薯丁等产品,在加工过程中需要搅拌机来混合各种原料。在研究和比较了现有的各种搅拌机的基础上,设计一种用于薯饼加工的搅拌机,实现了马铃薯薯饼的各种配料的混合。并通过多次试验检验该搅拌机的效果,混合均匀度在90%以上,满足薯饼加工工艺的需要。     

马铃薯品质无损检测技术研究进展

随着马铃薯产量的加大及其加工业的快速发展,马铃薯品质无损检测技术的研究对保障马铃薯加工和深加工业快速安全的发展起到了至关重要的作用。针对马铃薯内外品质的检测,包括大小、形状、颜色、病斑、腐烂、机械损伤等外部缺陷和马铃薯干物质含量、空心、黑色心腐等内部缺陷,分别从超声波检测技术、机器视觉检测技术、近红外检测技术、高光谱检测技术4个方面入手,并结合多种化学计量学方法及统计学方法分析国内外的研究现状,同时为满足马铃薯实时、高效的在线检测,提出建立完善的马铃薯高光谱检测技术理论,而在实际生产加工环节采用机器视觉与多光谱技术代替高光谱成像技术的新思路。     

矩形竹片动态缺棱检测装置设计与实验

论述了矩形竹片动态缺棱检测装置的设计研究,该检测设备由缺棱检测装置、颜色分类检测装置、进给装置、竹片收集装置组成。采用机械滚轮和杠杆结构配合拉线传感器实现矩形竹片缺棱检测。进行了竹条检测实验,分析实验结果得出,该装置可实现竹片缺棱自动化分选。     

海产品智能分拣装备与系统设计

目的：解决当前中国水产行业分选加工效能不足、准确率低、成本高的现状问题。方法：选用智能机器人与传统机械相结合开发新的机械结构对海产品进行分级;同时运用基于机器视觉的智能识别技术对海产品进行分类(采用了适应海产品的改进的新型Faster R-CNN算法),协同构建了一个能够智能、高效、自动分拣海产品的装备系统。结果：改进的新型Faster R-CNN算法识别技术在海产品检测中具有更高的检测准确率(海螺75.1%,扇贝80.3%,贻贝82.9%,蛤蜊86.2%)。结论：装备系统可以代替传统的人工操作模式,完成海产品的智能分拣过程。     

基于YOLOv5网络的缺陷硬糖智能分拣系统研究

针对糖果生产企业采用人工方法分选缺陷硬糖存在的漏检、成本高、效率低等问题,研究设计一种基于卷积神经网络的缺陷硬糖智能分选系统.通过工业相机采集硬糖图像,利用YOLOv5卷积神经网络模型进行缺陷硬糖的检测识别,使用喷阀剔除缺陷硬糖.测试结果表明,实时检测准确率高达98％,具有高度自动化和智能化水平,在食品生产工业中具有一...     

西北旱区不同品种马铃薯薯饼加工品质特性分析

为研究不同品种马铃薯与薯饼品质特性之间的关系,以适宜西北旱区栽培的13个马铃薯鲜薯为原料制备马铃薯薯饼,并对薯饼的感官和TPA进行了分析和比较,结果表明,不同马铃薯品种对加工薯饼的品质特性具有显著影响(p<0.05),不同品种马铃薯淀粉含量、蛋白质含量与马铃薯薯饼的硬度、弹性、胶着性及咀嚼性呈极显著正相关(p<0.01),而不同马铃薯品种淀粉含量、粗蛋白含量与马铃薯薯饼色泽无显著相关性(p>0.05);综合评价感官和TPA,供试的13个品种中,陇薯7号、陇薯8号、中薯18号和新品系LY08104-12的表现较好,制得的马铃薯薯饼感官品质较好,硬度适中,口感酥软,质构品质也较优。该结论可为不同马铃薯品种在薯制品加工适宜性方面提供一定的理论参考。     

基于模拟退火算法改进遗传算法的织物智能配色

为提高传统配色方法及现有配色算法的配色精度、效率及泛化能力,构建了基于BP神经网络的遗传算法和模拟退火算法相结合的织物智能配色模型,利用BP神经网络预测颜色,将训练好的BP神经网络与CIEDE2000色差公式结合作为遗传算法的适应度函数,用模拟退火算法改进的基于BP神经网络的遗传算法预测颜色配方,并根据预测的配方对涤纶织物进行染色实验,计算实验色差。结果表明:模拟退火算法优化的基于BP神经网络的遗传算法配色模型只需经过80次迭代即可收敛,预测颜色的理论色差均值为0.165,染色实验色差均值为0.289,配方绝对误差平均值为0.010 7;验证样本的理论色差均值为0.240,染色实验色差均值为0.437。该算法可实现织物的智能配色。     

硫色镰刀菌（Fusarium sulphureum）接种对抗病/易感品种马铃薯块茎苯丙烷代谢的影响比较

以抗病品种“青薯168”和易感品种“陇薯3号”马铃薯块茎为试材,比较2 个品种马铃薯块茎、切片接种硫色镰刀菌（F. sulphureum）后苯丙烷代谢关键酶活性和相关产物积累的动态变化。结果表明,“青薯168”马铃薯块茎及其切片接种F. sulphureum后的病斑直径显著小于“陇薯3号”。2 个品种马铃薯接种F. sulphureum后,在前期苯丙氨酸解氨酶（PAL）、4-香豆酰-辅酶A连接酶（4CL）活性以及总酚、类黄酮含量增加,但随着病斑的扩展,PAL、4CL活性以及总酚、类黄酮含量迅速降低并低于未接菌对照。肉桂酸-4-羟化酶（C4H）活性和木质素含量在整个培养过程中均显著低于未接菌对照。抗病品种“青薯168”接种F. sulphureum后,PAL、4CL和C4H活性以及总酚、类黄酮和木质素含量均显著高于“陇薯3号”。表明苯丙烷代谢在F. sulphureum与马铃薯块茎互作的早期发挥了积极的作用,一旦病斑开始扩展,苯丙烷代谢的抗病作用将显著降低;抗病块茎的苯丙烷代谢活性显著高于感病块茎,表明苯丙烷代谢在马铃薯块茎抗干腐病方面具有重要的作用。     

马铃薯鲜薯馒头的研制

以马铃薯鲜薯-小麦混合粉为原料制作馒头,研究了不同添加量的马铃薯鲜薯干物质、增筋剂(谷朊粉)和护色剂(Vc)对马铃薯鲜薯馒头加工特性的影响,并采用正交试验优化了马铃薯鲜薯馒头的加工工艺。结果表明,随着马铃薯鲜薯干物质添加量的增加,馒头的比容逐渐下降,咀嚼度逐渐增强,馒头的品质随之变差;增筋剂和护色剂的使用均可改善馒头的蒸煮品质;当马铃薯鲜薯干物质添加量为13%时,按混合粉质量(不包含马铃薯鲜薯所含水分的质量)分别添加70mg/kg Vc、1.5%谷朊粉和56%水,经传统加工工艺,可制得品质较好的马铃薯鲜薯馒头。     

马铃薯贮藏加工性能与开发利用

马铃薯是我国主要的粮菜兼用的作物，又是重要的工业原料，具有很高的开发利用价值，是新世纪我国最有发展前景的经济作物之一。马铃薯营养成分丰富而平衡，优于一般粮食和蔬菜，是营养健康食品之一，备受关注和青睐。种植马铃薯的投入产出比可达1：3～4。以马铃薯为原料可以开发出一系列的加工产品。将鲜薯加工成原粉、薯饼、薯泥，薯片等用于快餐或进入超市；淀粉原粉加工成系列变性淀粉，用于食品、纺织、医药、铸造、饲料等领域。我国是马铃薯种植面积最大的国家，2003年中国马铃薯种植面积达452．4万亩（hm^2），占世界25％，总产量约在6800万t。     

前处理和干燥对马铃薯全粉品质影响的研究进展

马铃薯全粉是新鲜马铃薯经去皮、切片、漂烫、护色等后脱水干燥制成的片状或粉末状产品,包含了除薯皮以外大部分马铃薯的营养物质,因此马铃薯全粉具有营养丰富、耐储的优点。作为马铃薯主粮化的原料,马铃薯全粉有助于改善和丰富我国居民膳食营养结构。近两年对马铃薯全粉的研究较多地集中在添加了马铃薯全粉制成的主食类或者其他相关产品的开发上。马铃薯全粉品质除了与马铃薯品种有关外还受到马铃薯全粉加工工艺的影响。本文将从马铃薯全粉生产过程中的前处理和干燥两个方面对马铃薯全粉品质的影响进行综述,以期为该领域的研究者提供理论参考。     

马铃薯生全粉制备过程中复合护色工艺的优化

以新鲜马铃薯为原料加工马铃薯生全粉,对加工工艺中的护色技术进行研究。通过考察柠檬酸、L-半胱氨酸和氯化钠3种护色剂对马铃薯生全粉色泽的影响,优选出3种护色剂合适添加范围;通过正交试验,得到柠檬酸-L-半胱氨酸-氯化钠无硫复合护色剂最佳护色配比为:0.5%柠檬酸、0.1%L-半胱氨酸、0.5%氯化钠。该复合护色优于单一护色效果,可使生全粉保持良好色泽,使制得马铃薯生全粉亮度值L^*达91.09,多酚氧化酶的抑制率87.98%。研究为马铃薯生全粉加工中的无硫复合护色工艺提供参考。     

微波消解-ICP-MS法测定红薯中18种元素

建立电感耦合等离子体质谱法和在线内标法测定3个红薯品种(黄心红薯、白心红薯、紫红薯)生红薯、熟红薯、红薯皮及表面土壤中铅(Pb)、砷(As)、镉(Cd)、汞(Hg)、铬(Cr)、钴(Co)、铝(Al)7种重金属元素及钠(Na)、钾(K)、钙(Ca)、锌(Zn)、硒(Se)、镁(Mg)、铁(Fe)、锰(Mn)、镍(Ni)、铜(Cu)、锡(Sn)11种微量元素的方法。对红薯样品微波消解体系进行优化,红薯样品经微波消解后,进行ICP-MS测定,并对方法的线性关系、检出限、定量限、回收率、精密度进行考察。进行3种生红薯、熟红薯、红薯表皮、红薯表面附着土壤的研究,用加标回收对测定方法准确度进行验证。结果表明,选择硝酸-双氧水的消解体系,样品消解更完全;18种元素标准品的标准曲线的相关系数R≥0.994,回收率86.8%~98.3%,重金属元素的方法检出限均低于0.5μg/L,微量元素的方法检出限均低于50μg/L;相对标准偏差小于5%;不同品种生红薯、熟红薯、红薯皮及表面土壤中重金属元素及微量元素含量各有差异。方法快速、准确、灵敏度高、定量准确,可用于红薯中重金属及微量元素的测定。     

Identifying Potato Varieties Using Machine Vision and Artificial Neural Networks

The objective of this study is to develop a method for identifying and discriminating 10 potato varieties by combining machine vision and artificial neural network methods. The potato varieties include Agria, Savalan, Florida, Fontaneh, Natasha, Verona, Karso, Elody, Satina, and Emrad. A total number of 72 characteristic parameters specifying color, textural, and morphological features are found among these varieties. By using principal component analysis, 16 principal features are selected for identifying and discriminating potato varieties. The data obtained from image processing were classified using linear discriminant analysis and non-linear artificial neural network method. The accuracy of discriminant analysis were 73.3, 93.3, 73.3, 40, 73.3, 73.3, 66.7, 80, 40, and 53.3%, respectively, for the varieties used in this study. The classification accuracy was improved by 100% for all the varieties using neural network analysis and the correct classification ratio was 100% using this method. It is revealed from the results that machine vision technique and neural network analysis could identify potato varieties with acceptable accuracy.     

基于轻量卷积网络的马铃薯外部缺陷无损分级

目前马铃薯外部缺陷检测方法主要依靠人工提取特征,且检测精度不高,为了更好地对马铃薯外部缺陷进行快速、准确在线分级,本实验提出一种基于轻量卷积网络的在线分级方法。首先,利用ImageNet数据集训练Xception网络模型,建立马铃薯预训练网络模型;然后,重新构建5 类缺陷全连接层,并通过迁移学习在预训练网络模型上训练马铃薯缺陷数据集;最后,基于外部缺陷识别模型分别测试5 类缺陷的分级性能。结果表明,学习率为0.000 01时,网络模型整体性能最优,训练准确率为98.88%,损失值为0.034 9;在相同样本条件下,与9 种不同深度的网络进行对比,本实验构建的轻量级网络模型识别效果最好,平均识别准确率达到96.04%,且运行时间比识别效果较好的ResNet152网络更短,本实验网络模型的识别速率为6.4 幅/s,本研究结果可为马铃薯在线分级提供理论支持。     

TIMED COLOR CHANGES AMONG RAW AND BAKED POTATO VARIETIES

Color changes in twenty two lines of potatoes by the browning reaction were measured and evaluated by Hunter color value and National Bureau of Standard (NBS) unit of color difference (ΔE). The relationship between color change and total solids content was evaluated (p < 0.05). CO87106–5 and AC87313–3 contained the highest solids content, while CO86218–2 and CO82142–4 had the lowest solids content. Overall color change in raw potatoes by browning showed CO86142–3 with the least color change and AC87084–2 with the most. Hunter color values for the browning reaction changed with time. As time increased, potatoes became darker, redder and bluer. Moreover, all of the Hunter color values were significantly changed by baking. Color of baked BC0894–2, CO86142–3 and CO87106–5 were the least changed, while that of the baked CO8001–5, Russet Norkotah and TC1406–1 were the most changed. Degree of browning in raw potato was not correlated with total solids content, but did show a negative relationship (r =−0.12). Also, rate of color change by the browning reaction in raw potato was not related to potato total solids (r =−0.10). Color change by baking also was not correlated with total solids content (r =−0.36). Raw potato color change was positively correlated to baked potato color change (r = 0.42). Genotype of the potatoes was closely related with     

不同加工方式对马铃薯原粉特性的影响

以新鲜马铃薯为原料,制备得到马铃薯熟粉、冷冻干燥马铃薯原粉、未经烫漂的热风干燥马铃薯原粉及经烫漂的热风干燥马铃薯原粉,研究不同类别的马铃薯原粉在外观品质、理化特性、功能特性及复水特性等方面的性质。结果表明,不同加工方式下马铃薯原粉的基本理化性质及外观品质均存在显著差异,其中经烫漂的热风干燥马铃薯原粉色泽(L*值为80.78)最佳,为淡黄色粉末,具有高淀粉、高纤维及低脂肪的特点,有利于用作减肥及降血脂健康食品的原料;功能特性结果表明,马铃薯熟粉具有较高的持水能力(3.03±0.01 g/g)、乳化性(21.58%±1.37%)及乳化稳定性(15.98%±0.73%);冷冻干燥马铃薯原粉具有较高的持油能力(3.42±0.08 g/g)和溶解性(16.49%±1.07%);经烫漂的热风干燥马铃薯原粉具有较高的乳化稳定性(16.67%±0.76%)、复水性(复水比为6.43)和复水稳定性(95.76%±0.85%)。同时,结合考察不同加工方式下马铃薯原粉的加工效率与能耗,结果发现,经烫漂的热风干燥原粉具有较高的加工品质且加工能耗低,可作为熟粉的替代品。本研究可为马铃薯不同产品特点选择合适的加工方式,并为马铃薯原粉的研制及代替熟粉的可行性奠定基础。     

The influence of French fries processing on the glycoalkaloid content in coloured-fleshed potatoes

French fries made from coloured-fleshed potatoes may be interesting alternative to the traditional snacks for consumers. However, potatoes contain glycoalkaloids (TGA), so potato tubers and obtained fried snacks should be subjected to comprehensive examination. The aim of this study was to determine the effect of different stages of French fries processing on the content of TGA (α-solanine and α-chaconine) in the red- and blue-fleshed potatoes, in semi-products and ready-to-eat products. It was stated that during the processing of French fries prepared from coloured-flesh potato varieties significantly decreased the content of TGA (α-solanine and α-chaconine) in the samples obtained at different stages of the process compared to the raw material. Potatoes with blue-fleshed of Vitelotte variety and red-fleshed of Highland Burgundy Red variety can be used to French fries processing due to their low content of TGA (in unpeeled and peeled potatoes). However, Blue Congo variety with blue-fleshed should not be applied to French fries processing, because of high TGA content in raw material and first of all in peeled potatoes flesh. The peeling process of coloured-fleshed potatoes decreased in TGA content on average by about 50 %, cutting process on average by about 53 %, whereas blanching on average by about 58 % compared with the raw material. The highest decrease in TGA content was caused by frying process. The mean values were about 97.5 % in ready-to-eat French fries. In French fries after I and II steps of frying, the ratio α-solanine to α-chaconine was lower (1.0:2.0) than in unpeeled potatoes (1.0:2.3).     

基于机器视觉的淡水鱼去鳞损伤无损检测识别方法

为了研究淡水鱼去鳞损伤无损检识别方法,以鲤鱼为研究对象,搭建视觉采集系统,实现图像采集与处理。通过采集鲤鱼去鳞后有去鳞损伤的图像,将样本图像划分为鱼腹、鱼背、鱼尾、损伤、背景五部分,提取感兴趣区域(ROI),比较其颜色差异,发现正常鱼体的表面区域与损伤区域颜色差异较大,以不同区域对应的红色通道值(R)、绿色通道值(G)、蓝色通道值(B)对应像素的平均值作为输入,建立基于广义神经网络(GRNN)和径向基函数神经网络(RBF)的鱼体损伤区域检测识别模型,经测试,识别准确率分别能达到98%和80%。通过计算预测损伤区域与实际损伤区域的相关面积大小,验证识别模型的准确度,经试验验证,基于GRNN和RBF网络识别模型对去鳞损伤预测的准确率分别达到了90%和83%。研究结果表明,机器视觉可应用于淡水鱼去鳞损伤的无损检测识别。     

A Deep Convolutional Neural Network Architecture for Boosting Image Discrimination Accuracy of Rice Species

In this paper, we proposed a machine vision system based on deep convolutional neural network (DCNN) architecture for improving the accuracy of classifying three distinct groups of rice kernel images compared with the traditional approaches. The main advantage of the presented method was able to avoid many heuristics as well as manual labor to tune complex parameters according to the domain to reach a modest level of accuracy in the classical feature extraction algorithms. We trained our models using stochastic gradient descent with momentum of 0.9 and weight decay of 0.0005 to optimize the network parameters and minimize the back-propagation error on the training dataset. We used a batch size between 15 and 150 and epochs time configured between 10 and 25. The experiment results showed that the highest accuracy of 99.4% obtained in the training process with batch size of 15 and epoch time of 20. We also compared the DCNN method with the traditional hand-engineered approaches of PHOG-KNN, PHOG-SVM, GIST-KNN, and GIST-SVM for rice kernel classification. The results showed that DCNN routinely outperforms other methods in similar machine vision tasks. The prediction accuracy results for test datasets by PHOG-KNN, PHOG-SVM, GIST-KNN, and GIST-SVM models were 89.1, 76.9, 90.6, and 92.1%, respectively. The highest prediction accuracy of DCNN is 95.5%, which showed the effectiveness of our proposed method for rice kernel classification. The aim of this study is to set up an automatic and accurate intelligent detection system and offering much value to current rice processing industry. With the comparably high classification accuracy, developed neural network could be used as a tool to achieve better and more objective rice quality evaluation at trading points within the rice marketing system.