| 查看: 1108 | 回复: 1 | |||
[求助]
摘要润色
|
|
中国是世界上最大的苹果生产国和消费国，苹果种植面积和产量均占世界总量的40%以上，在世界苹果产业中占有重要地位。而苹果在采摘或者运输过程中，会不可避免的因为一些外力原因造成不同程度的损伤，有些损伤表面看不出来或者不明显，但是损伤部位的内部品质却已经发生变化，所以对于苹果外部损伤的检测至关重要，同时消费者在挑选水果时从以往仅仅关心水果的外部品质，也开始注重水果的内部品质。 高光谱成像技术将传统的二维成像技术和光谱技术有机结合，可同时获得被测物体的图像信息和光谱信息，本文以糖心富士苹果为研究对象，利用高光谱成像技术（380-1038nm）对苹果的外部损伤、内部品质（糖度、pH值）、苹果口感指标（酸味、涩味的回味）进行了研究，主要内容和结论如下： （1）利用高光谱成像技术对苹果外部损伤进行检测，首先利用高光谱成像系统获取可见-近红外波段（380nm-1038nm）的苹果图像，对全波段苹果图像做一次主成分分析，选取每个样本最能区分损伤区域和正常区域的主成分图像，其次根据该主成分图像的特征向量优选出10个特征波段，针对特征波段再做一次主成分图像，选取第四个主成分图像（PC-4）做图像处理和识别，其综合识别正确率只有81%。原因是由于苹果特性，采集的苹果数据会存在光斑从而影响实验效果，为消除光斑的影响，选取了合适的基准图像进行图像处理，利用图像差值算法，将综合识别率提高到90%。 （2）利用高光谱技术对苹果内部糖度和pH值进行无损检测研究，通过对原始光谱分别进行了多元散射校正(MSC)、Savitzky-Golay(S-G)卷积平滑和MSC+S-G的光谱预处理，并利用偏最小二乘回归(PLSR)和主成分回归(PCR)对全波段光谱建模。对比不同的光谱预处理后的模型效果，优选出糖度值和pH值光谱的特征波段，并对特征波段进行建模分析。综合各方面的因素分析，特征波长下建立的模型优于全波段下建立的模型。 （3）结合电子舌技术和高光谱技术检测苹果的酸味和涩味的回味两个口感指标，通过对原始光谱分别进行了多元散射校正(MSC)、Savitzky-Golay(S-G)卷积平滑和MSC+S-G的光谱预处理，并利用偏最小二乘回归(PLSR)光谱建模。对比不同的光谱预处理建模效果，对酸味进行PLSR建模效果最好的是MSC预处理方法，对涩味的回味进行PLSR建模效果最好的是MSC+S-G预处理方法。 China is the largest producer and consumer of apples in the world, Apple growing area and production accounted for more than 40% of the world's total, and china plays an important role in the world's apple industry. In the process of picking or transportation, Apple will cause different degrees of damage because of some external forces., Some of the damaged surface can not see or not obvious, but the internal quality of the damaged site has changed, so it is important to detect the external damage of apple. At the same time consumers not only care about the external quality/ of fruit, but also began to pay attention to the internal quality of fruit. Hyperspectral imaging technology combines the traditional two dimensional imaging technology and spectral technology, which can obtain the image information and spectral information of the measured object simultaneously. The apple named sugar heart Fuji as the research object, using hyperspectral imaging technology (380-1038nm) on Apple's external injury, internal quality (Brix value, pH value), apple taste (sour and astringent aftertaste) was studied. The main contents and conclusions are as follows: (1) Using hyperspectral imaging technology to detect the external damage of apple. Firstly, hyperspectral imaging system was used to acquire the image of the visible near infrared band (380nm-1038nm),Then through the principal component analysis method to detect the whole band apple image and select the principal component image of each sample which can Distinguish between damage area and normal area easily. Secondly, according to the feature vector of the principal component image, the 10 characteristic bands are selected, and then the principal component analysis is done for the characteristic band, the fourth principal component images (PC-4) were selected to do the image processing and recognition, and the recognition accuracy was only 81%. The reason is due to the characteristics of the apple, the apple data will exist in the spot so as to affect the experimental results. In order to eliminate the influence of the spot, a suitable reference image is selected for image processing, and the recognition rate is increased to 90% by using the image difference algorithm. (2) Using hyperspectral technology Nondestructive to detect the sugar content and pH value of apple.The algorithm of multiplicative scatter correction spectral pretreatment (MSc), savitzky Golay (S-G) convolution smoothing and MSC+S-G were used to deal with the original spectrum, then partial least squares regression (PLSR) and principal component regression (PCR) were used to model the full band spectral modeling. Compared with the different spectral pretreatment model, the characteristic bands of Brix value and pH value were optimized, and the characteristic band was modeled and analyzed. Based on the analysis of various factors, the model established under the characteristic wavelength is better than that of the whole wave band. (3) Combined with electronic tongue and Hyperspectral technology to detect apple’s taste index of sourness and astringent aftertaste. The algorithm of multiplicative scatter correction spectral pretreatment (MSc), savitzky Golay (S-G) convolution smoothing and MSC+S-G were used to deal with the original spectrum, then partial least squares regression (PLSR) was used to model the full band spectral modeling. Comparison of different spectral pretreatment modeling results, MSC pretreatment method for the PLSR modeling of sour taste was the best,while MSC+S-G pretreatment method for the PLSR modeling of astringent aftertaste was the best. The results show that the use of hyperspectral imaging technology can effectively detect the quality of Apple's internal and external, and it provided experimental basis for the subsequent online detection. |
回复此楼» 猜你喜欢
A期刊撤稿
已经有5人回复
-
临港实验室与上科大联培博士招生1名
已经有8人回复
-
26申博自荐
已经有7人回复
-
想换工作。大多数高校都是 评职称时 认可5年内在原单位取得的成果吗？
已经有4人回复
-
带资进组求博导收留
已经有9人回复
-
求助大佬们，伤口沾上了乙腈
已经有6人回复
-
最近几年招的学生写论文不引自己组发的文章
已经有9人回复
【答案】应助回帖
| China is world's largest apple producer and consumer whose apple growing size and production accounted for more than 40% of the world's total, therefore China plays an important role in the world's apple industry. In the process of picking and transportation,different degrees of damage will be done to apples because of some external factors.Some of the damaged surfaces are not obvious, but the internal quality under the damaged parts has been changed, so it is important to detect the external damage. While consumers pay attention to the appearance of fruit, they value the internal quality of fruit more nowadays. |
2楼2016-03-18 17:57:04