CN111624265A

CN111624265A - Method for identifying egg species

Info

Publication number: CN111624265A
Application number: CN202010320172.2A
Authority: CN
Inventors: 韩敏义; 石金明; 邓绍林; 王晓明; 郑桂青; 徐幸莲; 周光宏; 黄启荣
Original assignee: Guangdong Wens Jiawei Food Co ltd; Nanjing Agricultural University; Wens Foodstuff Group Co Ltd
Current assignee: Guangdong Wens Jiawei Food Co ltd; Nanjing Agricultural University; Wens Foodstuff Group Co Ltd
Priority date: 2020-04-22
Filing date: 2020-04-22
Publication date: 2020-09-04

Abstract

The invention provides a method for identifying egg types. Detecting the eggs of known varieties by adopting gas chromatography-ion mobility spectrometry to obtain GC-IMS sample data; establishing a new model by using LAV software, adding sample data into an engineering file manager, carrying out color designation for distinguishing, and integrating sample peaks; deriving the peak volume data for the generated sample into a comma separated file; importing a machine learning related library into python software for data cleaning; then, segmenting the data; establishing a random forest classification model; and detecting the eggs of unknown varieties to be detected by using gas chromatography-ion mobility spectrometry, and applying the obtained sample data to the established random forest model to obtain the varieties of the eggs to be detected. The method can accurately test the variety of the eggs, does not need to preprocess the samples, is simple and convenient to operate, saves labor force and saves detection time.

Description

Method for identifying egg species

Technical Field

The invention relates to the field of agricultural product detection, in particular to an egg type identification method.

Background

Eggs are one of the best nutritional sources for humans, and contain high biological value proteins and a wide variety of vitamins and minerals. For humans, egg proteins are second only to breast milk. The quality of eggs is directly related to the personal interest of the consumer. The contents of protein and lecithin in the primary egg are higher than those in the common egg, and the lecithin is the best substance for promoting hepatocyte regeneration, promoting brain development, enhancing memory and supplementing brain nutrition. The primary egg is produced from the primary hen within 30-60 days before the primary hen starts laying, and is the essence of the primary hen essence. The weight of the primarily laid eggs is generally 35-45 g. Each laying hen can lay a small primary egg by accumulating 3-9 days of nutrition on average. Thick and thin, dark yolk, thick egg white, no hormone and antibiotic, low cholesterol, delicious taste, rich vitamins, amino acids and various trace elements needed by human body, and the nutrient content of the first-laying eggs with equal weight is relatively higher compared with the common eggs. Therefore, the price of the primary egg is several times of that of the common egg. Some common household eggs are sold as newborn eggs, which not only disturbs the market, but also damages the interests of consumers. Therefore, a detection technology is needed to be researched to identify the egg types, and the method has important significance for distinguishing good eggs from bad eggs, striking 'good order' bad merchants and maintaining the market fairness.

Disclosure of Invention

In view of the above, it is desirable to provide a method for identifying egg types.

In order to achieve the purpose, the invention adopts the following technical scheme:

the method for identifying the egg type comprises the following steps:

step S1: detecting the eggs of known varieties by adopting gas chromatography-ion mobility spectrometry to obtain GC-IMS sample data;

step S2: establishing a new model by using LAV software, adding sample data into an engineering file manager, carrying out color designation for distinguishing, and integrating sample peaks;

step S3: deriving the peak volume data for the generated sample into a comma separated file;

step S4: importing a machine learning related library into python software for data cleaning; then, segmenting the data; wherein 2/3 is a training set, 1/3 is a verification set, the training set is used for establishing a random forest classification model, and the verification set is used for evaluating the classification accuracy of the model;

step S5: and (4) detecting the eggs of unknown varieties to be detected in the steps S1-S3, and applying the obtained sample data to a random forest model to obtain the varieties of the eggs to be detected.

Further, in the above method for discriminating the kind of egg, the discrimination of the egg is the discrimination of a primary egg and a normal egg.

Further, in the above method for discriminating a kind of an egg, the step of detecting an egg in step S1 includes:

step 1: removing shell of egg, placing into beaker, and stirring with glass rod;

step 2: putting 1-2 mL of the uniformly stirred egg liquid into a headspace bottle;

and step 3: putting the headspace bottle into a sample rack, and incubating for 10-20 min;

and 4, step 4: injecting sample by using an injector of an automatic sample injector, wherein the sample injection volume is 0.2-0.6 mL;

and 5: and (3) carrying out analysis by GC-IMS, wherein the carrier gas is nitrogen, EPC1 is 150mL/min, EPC2 is 2mL/min, the duration is 5min, and then EPC2 is kept at 50-80 mL/min.

Further, in the method for identifying the type of the egg, the incubation temperature in the step 3 is 40 to 50 ℃.

Further, in the above method for identifying the type of egg, the peak volume above the baseline is selected as the intensity type using the new model created by the LAV software in step S2, and the correction function type is boltzmann.

Further, in the above-described method for discriminating the kind of egg, the machine learning-related libraries in step S4 are pandas and sklern.

According to the method for identifying the egg variety, the sample to be detected is detected by using the gas phase-ion mobility spectrometry, and the sample does not need to be subjected to complex pretreatment before detection, so that a large amount of labor force is saved. The detection efficiency is effectively improved, the detection time is about 30min, and the detection time is shortened. In addition, the random forest classification model established by the method is simple and convenient, factors such as marker content do not need to be compared, and the identification accuracy is high.

Drawings

FIG. 1 is a schematic flow diagram of the present invention;

FIG. 2 is a random forest classification model 1 st tree;

FIG. 3 is a 50 th tree of the random forest classification model;

FIG. 4 is a 100 th tree of the random forest classification model;

FIG. 5 is a confusion matrix for predicting eggs of unknown breeds.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions of the present invention will be further clearly and completely described below with reference to the embodiments of the present invention. It should be noted that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Example 1

Taking 20 eggs of known varieties, including 10 newborn eggs and 10 common eggs, and establishing a random forest classification model according to the following steps:

(1) after 20 eggs are shelled, respectively putting the eggs into different beakers, and uniformly stirring the eggs by a glass rod; respectively putting 1mL of the uniformly stirred egg liquid into different headspace bottles; putting the headspace bottles into a sample rack in sequence, and incubating for 10 min; the incubation temperature was 50 ℃; injecting sample by using an injector of an automatic sample injector, wherein the sample injection volume is 0.5 mL; performing analysis by GC-IMS, wherein the carrier gas is nitrogen, EPC1 is 150mL/min, EPC2 is 2mL/min, the duration is 5min, and then EPC2 is kept at 50 mL/min;

(2) establishing a new model by using LAV software, selecting the peak volume above a base line according to the intensity type, and setting the correction function type as Boltzmann; adding sample data into an engineering file manager, carrying out color designation for distinguishing, and integrating sample peaks;

(3) deriving the peak volume data for the generated sample into a comma separated file;

(4) introducing pandas and sklern into python software for data cleaning; then, segmenting the data; wherein 2/3 is a training set, 1/3 is a verification set, the training set is used for establishing a random forest classification model, and the verification set is used for evaluating the classification accuracy of the model. The parameters set when the random classification model is established are as follows: impurity rating scale: a coefficient of kini; maximum depth of tree: none; maximum number of features: automatic; maximum depth: none; minimum number of samples required to split an internal node: 2; minimum number of samples required for leaf node: 1; the number of classification trees is as follows: 100, respectively; the generalization accuracy was estimated for the out-of-bag samples: if not; a random number generator: 222.

the generated random classification model is composed of 100 classification trees. Fig. 2, 3, 4 are the 1 st, 50 th and 100 th trees of the random forest classification model, respectively.

Example 2

Taking 40 egg samples of unknown varieties, wherein the number of eggs of different varieties is known by an egg sample taking person, and the egg samples are sent to a tester for testing, and the tester performs testing according to the following steps:

(1) removing shells of eggs, respectively placing the eggs into different beakers, and uniformly stirring the eggs by using a glass rod; respectively putting 1mL of the uniformly stirred egg liquid into different headspace bottles; putting the headspace bottles into a sample rack in sequence, and incubating for 10 min; the incubation temperature was 50 ℃; injecting sample by using an injector of an automatic sample injector, wherein the sample injection volume is 0.5 mL; performing analysis by GC-IMS, wherein the carrier gas is nitrogen, EPC1 is 150mL/min, EPC2 is 2mL/min, the duration is 5min, and then EPC2 is kept at 50 mL/min;

(4) the sample data was applied to the generated random forest model in example 1 to predict egg type of unknown breed.

The predicted results are shown as a confusion matrix in fig. 5. As can be seen from fig. 5, the accuracy of the detection of the established random forest model in example 1 is 100%.

The above-mentioned embodiments only express several embodiments of the present invention, and the description thereof is more specific and detailed, but not construed as limiting the scope of the present invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the inventive concept, which falls within the scope of the present invention. Therefore, the protection scope of the present patent shall be subject to the appended claims.

Claims

1. The method for identifying the egg type is characterized by comprising the following steps:

step S3: deriving peak volume data for the generated samples, comma separated files;

step S5: and (4) detecting the eggs of unknown varieties to be detected in the steps S1-S3, and applying the obtained sample data to the established random forest model to obtain the varieties of the eggs to be detected.

2. The method for discriminating egg types according to claim 1, wherein said discrimination of eggs is discrimination of a virgin egg and a normal egg.

3. The method for discriminating egg kind according to any one of claims 1 or 2, wherein the step of detecting the egg in step S1 includes:

4. The method for identifying the kind of egg according to claim 3, wherein the incubation temperature in step 3 is 40 to 50 ℃.

5. The method for discriminating an egg kind according to claim 1, wherein the intensity type is selected as a peak volume above a base line and the correction function type is boltzmann using a new model created by LAV software in step S2.

6. The method for discriminating egg kind according to claim 1, wherein the machine learning related library in the step S4 is pandas and sklern.