A Preliminary Study On The Identification Of Transgenic Crops Based On Terahertz Time Domain Spectroscopy And Imaging Technology

The potential risk of genetically modified crops has been the focus of attention,and it's testing technology research is the primary task and key to the identification and supervision of transgenic crops.Traditional transgenic detection methods are cumbersome,costly and long cycle,and are destructive to the research object which could not meet the requirements of real-time monitoring in the current agricultural production process.Therefore,the development of a non-destructive,fast and accurate transgenic detection technology has become an important work in agricultural production.Terahertz is a frequency of 0.1-10 THz electromagnetic waves,with low energy,coherence and fingerprint spectrum,etc.,the detection of protein,DNA,amino acids and other biological macromolecules have achieved good results,due to the broad application prospects in the field of non-destructive testing,it has become the hot areas of scientific research this century.In this paper,terahertz time domain spectroscopy and imaging techniques were used to identify the hot genetically modified crops(transgenic soybean and transgenic rice).The main findings of the study are as follows:1)The terahertz time domain spectra of transgenic soybean,common soybean and hybrid soybean were obtained using the terahertz time domain spectroscopy system.The frequency domain spectrum of the sample was obtained by fast Fourier transform.Based on the frequency domain spectrum and principal component extraction,the particle size optimization algorithm is used to solve the parameter optimization problem of LS-SVM.The experimental results showed that the proposed model is optimal and the correctness of the identification is 95%.2)According to the Fresnel reflection law,acquiring the reflection or refraction coefficient of the terahertz time domain spectrum on the surface and inner surface of soybean seed coat,the terahertz imaging of soybean seed reflection intensity and seed coat thickness was completed by visual imaging to realize the identification of transgenic soybean.3)Combining the terahertz images of rice seeds to achieve regional spectral signal extraction.Aiming at the problem that the main components of conventional rice and BT transgenic rice were not obvious,the K-SVD algorithm was used to construct the super-complete dictionary of terahertz spectral data of transgenic rice and its parents,and the sparse coefficient was solved by orthogonal matching tracing(OMP).The random forest algorithm of multi-feature sparse coefficient was constructed to improve the generalization ability and classification effect.