Research On The Classification Of Rice Species And The Chalkiness Test

Rice is one of the most important food crops in the world.The varieties and quality of rice are different,and there is a big gap in taste,nutrition and price.However,at present,the classification and chalkiness detection of rice is still in the primary stage,which is inefficient and inaccurate.Therefore,it is of great significance to develop an accurate,rapid,non-destructive,large-scale rice classification and chalkiness detection technology.Hyperspectral image can obtain the image information and spectral information of samples at the same time,which has great potential in the field of rice classification and chalkiness detection.In this paper,based on hyperspectral image technology,rice variety classification and rice chalkiness detection using 3D convolution neural network are studied,and compared with the methods based on support vector machine and convolution neural network.In the problem of rice classification,this paper first through hyperspectral image acquisition,MSC,SNV,Ma three spectral preprocessing,and then use lasso regression model for feature band selection,the processed samples into the model for classification.The results show that 3D convolutional neural network as a hyperspectral image data classification model has great advantages,the accuracy is 99.4%.In the problem of rice chalkiness detection,hyperspectral data are processed by similar methods,and input to support vector machine,convolution neural network and 3D convolution neural network for comparison.The results show that the accuracy of 3D convolution neural network is 98.5%,and the effect is the best.In a word,using hyperspectral image combined with 3D convolutional neural network can achieve pixel level rice classification and chalkiness detection accurately.The research of this technology can further promote the development of rice identification and quality detection technology to the direction of automation and intelligence,which has very important scientific significance and good practical prospects.