| The geographical indication rice(landmark rice)is rich in nutrients,sweet and delicious,and has been favored by consumers.The price of the landmark rice on the market is much higher than that of ordinary rice.At present,the traceability of the origin of the landmark rice mainly includes two parts: process traceability and source traceability.Process traceability may cause false origin and substandard phenomenon due to the influence of human factors,which seriously interferes with market order and damages the legitimate rights and interests of consumers and formal landmark rice enterprises(the state authorized to sell landmark rice).Therefore,this paper wants to establish a platform for the confirmation of the origin of rice,to quickly and accurately identify the origin of rice,and to trace the source.In this paper,a near-infrared spectrum image of 100 sets of rice samples from Liuhe landmark rice and Songyuan rice was collected,and the spectral data of rice was preprocessed.Principal component analysis was used to classify the data.However,due to the similar nature of rice,the principal component score map The overlap is serious,and the production area is not effective.Therefore,it is necessary to integrate the Support Vector Machines(SVM)algorithm to analyze the fingerprint and establish the landmark rice confirmation model.With this model as the core,by analyzing the needs of regular landmark rice enterprises and consumers,we design and implement a corporate rice fingerprint data platform based on Python language and Django framework,in order to easily identify the origin of rice and maintain the authority of landmark rice.Sex,safeguarding the fundamental interests of businesses and consumers.The main contents and results of this paper are as follows:(1)By comparing the influence of four different preprocessing methods on the discriminant rate of support vector machine model,it is determined that multivariate scatter correction is the best pretreatment method.(2)Selecting the radial basis kernel function as the kernel function of the support vector machine.When the penalty parameter c=4 and the kernel function parameter g=0.25,the rice vector confirmation model based on the support vector machine has the best discriminant effect,and the discriminant rate is reached.94%.(3)According to the platform requirements analysis,a total of eight modules,such as user login module,background management module,product display module and origin confirmation module,were designed.(4)Test the platform,the page test is good,the function test is normal,the user can directly submit the near-infrared spectrum detection data through the platform,and visualize the model test result,which can confirm the rice production. |
