Analysis and Research on the Characteristics of Boiled Yolk based on Hyperspectral Remote Sensing Images

Eggs are important nutritional foods in people's daily life. They are rich in protein and vitamins, as well as a variety of minerals like iron, calcium and phosphorus. Therefore, monitoring egg quality is always a focus problem that people are concerned about. This article breaks through traditional ideas and methods, proposes using hyperspectral remote sensing images to obtain the spectral characteristics of boiled yolk, and establishes the analysis model of egg quality. Through a large number of experiments on different types of eggs, the experimental results are effective, and this shows that the method proposed in this paper is feasible. It provides new ideas and methods for the examination of egg quality and the establishment of traceability system.


Introduction
Eggs are rich in protein, multivitamins, minerals and low fat, so they have become an important food substance in the world.Similarly, for the residents of our country, eggs have become the daily food of ordinary people.According to relevant information, in recent years, the total output of eggs in China has basically maintained around 23 million tons, accounting for 43% of the total egg output of the world, and the average per capita consumption is about 17kg, ranking the highest in the world.Therefore, the problem of the safety and quality of egg food is being paid more and more attention.There are often some food safety incidents in our daily life, such as the report of the European Union in 12 August 2017 that there are 16 countries involved selling poisonous eggs [1].
For the identification and safety supervision of egg quality, it mainly relies on the detection and identification techniques of eggs.The traditional identification techniques of egg quality include sensory identification and light identification.People use the transmission characteristics of eggs to detect and classify the freshness of eggs.With the development of electronic technology and the progress of science and technology, later a variety of testing instruments are developed and are used to obtain the electrical properties, acoustic properties, visual of the machine and optical characteristics of eggs, through processing to the obtained characteristics data, so as to realize the nondestructive detection of egg quality.Therefore, the nondestructive detection of egg quality has become the mainstream detection method.These methods include fluorescence spectrum [2], infrared spectrum [3,4], image processing [5,6], dielectric property detection [7], electronic nose [8] and hyperspectral images [9].Near infrared spectroscopy is also a very important method to detect eggs.Near infrared spectroscopy of eggs is used to establish a mathematical relationship model between egg spectral characteristics and quality indicators, and then it can effectively identify the quality of eggs.
Hyperspectral imaging realizes the combination of image and spectrum, so it can obtain the spectral characteristics of the material from the angle of the surface.It is obvious that the information obtained by hyperspectral remote sensing images is not exactly the same as the information obtained by the spectrometer, because the spectrometer obtains the spectral characteristics of the point, and the hyperspectral imager obtains the spectral characteristics of the surface.Therefore, this paper uses hyperspectral images to study the spectral characteristics of egg yolk and identifies the quality of eggs.This article breaks through the traditional way of thinking, and uses new techniques and methods to study the quality of eggs.The problem of egg quality detection is discussed from the spectral characteristics of boiled yolk in this paper, as has important reference value for the identification and traceability of egg quality.

Hyperspectral data collection
The comparison experiments of six kinds of different type eggs were carried out.Eggs of from the first to fourth kind are breeding eggs, the fifth is ordinary eggs, and the sixth is free range eggs.Five eggs are randomly selected from every type egg and they perform experiments.These selected eggs are washed, numbered and put into a pot, cooking for about 20 minutes.Take the cooked eggs out of the pot, we will perform the next experiments when the boiled eggs are cooled.
The computer and hyperspectral imager are started, and the corresponding imaging parameters are set up.After the shells of the cooked eggs are removed and the boiled egg white is broken, the boiled egg yolk is got.Put the yolk of each type of egg into a plate, use the hyperspectral imager to carry out scan imaging on egg yolk, and the hyperspectral image data of the boiled yolk is obtained.With the ENVI software to open hyperspectral image data, the spectral characteristic curve of egg yolk will be obtained.On each egg yolk, the spectral properties of several measuring points are obtained, and then they are saved.The spectral data is filtered and the size of the filter window is set to three.The data acquisition process is shown in Figure 1, and the obtained experimental data is shown in Figure 2 to Figure 7.

Data processing and analysis
It can be seen from the above given data of the spectral characteristics of boiled yolk that there are some abnormal spectral characteristic curves in the first and fifth kinds of eggs, so these anomalous characteristic curves are removed or processed.As we can see from Figure 2 to Figure 7, there is little difference in all curves in the red light part of the visible light.That is to say, the quality of eggs cannot be identified in this part of the visible light.However, there are some different curves between the green and the blue part of the visible light and the near infrared band.The biggest difference is the first type of egg, which shows that the storage time of the eggs selected in the first type of egg is quite different.In order to further analyse and compare the differences between different types of eggs, the obtained spectral characteristics are processed.A simple sum and normalization of all spectral characteristics of each type of egg is carried out.Namely, the average spectral characteristic curve of each type of egg is obtained, as shown in Figure 8.In Figure 8, we can see that the middle red light region cannot be identified, while the blue and near infrared regions at both ends can distinguish the freshness of different eggs.The greater the reflection value of the spectral characteristics is, the longer the time it is stored.In order to make it easy to distinguish, the spectral data are reordered and the reflected values are sorted from small to large, as shown in Figure 9.In Figure 9, although the abscissa is still expressed in wavelengths, it is not the actual wavelength position.It is very clear in Figure 9 that the average storage time of the fourth type egg is the longest, next is the third and the fifth type egg, followed by the first and the sixth type egg.The shortest storage time and the freshest are the second categories.It is the second type egg whose storage time is the shortest and whose freshness is the best.From the above analysis, we can draw a conclusion that the spectral characteristics of boiled yolk based on hyperspectral remote sensing images can only distinguish the freshness of eggs, and cannot identify different types of eggs.

Conclusion
It is an important way to ensure the quality of eggs and identify the fake and inferior products by exploring new methods of identifying the quality of eggs.In this paper, the hyperspectral image data of boiled yolk was obtained by the hyperspectral imager, and the spectral characteristics was extracted.By comparing the spectral characteristics of different types of egg yolk, we can see that hyperspectral images acquire data from the angle of the surface.Therefore, using hyperspectral remote sensing images to study the characteristics of boiled eggs, we can only recognize the freshness of eggs and not recognize the types.The results have a certain reference value for testing the freshness of eggs.

Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7
Figure 1 Chart diagram of acquiring spectral characteristics data of boiled yolk

Figure 8 Figure 9
Figure 8  The spectral characteristic curve of boiled yolk of different type eggs