Near-Infrared Spectral Modeling And Identification System Of Maize Haploid Seeds Maize Haploid Seeds

Haploid seed breeding technology is a major research topic of model agriculture.The haploid of crops can be used to obtain pure varieties of crops,and can also be used in hybrid breeding of crops to obtain hybrid varieties with excellent characteristics.Maize is a food crop with the largest planting scale and the highest yield in China.Haploid breeding is a fast and efficient way in maize breeding.Haploid identification is a prerequisite and necessary condition for the research and implementation of maize haploid breeding technology.The identification of maize haploid mainly depends on cell and anatomical identification,molecular marker identification and genetic marker identification,but these methods are not only time-consuming and laborious,but also cause damage to maize seeds.In order to help researchers to solve the above problems in distinguishing between haploid and diploid maize seeds,this paper used the advantages of near-infrared spectrum in identifying speed and accuracy of maize seeds induced by haploid,and developed a near-infrared spectral modeling and identification system of maize haploid seeds.For high oil induced maize seeds,this system can also assist in identification by predicting the oil content of the seeds to improve the accuracy of identification.This system uses Py Qt as the development basis of the system’s graphical interface,combined with Sklearn,Py Qt Graph and other extension libraries,and uses related algorithms such as Support Vector Machines,Partial Least Squares Regression,Multiple Scattering Correction and Principal Component Analysis in near-infrared spectrum technology.It realizes the functions of data arrangement,spectrum preprocessing,feature dimensionality reduction,establishment of identification model and sample identification for the original spectrum.And the specific maize seeds spectral data is used to verify the effectiveness of the function in this system.Researchers can use this system to establish maize haploid seeds identification models and find the optimal identification model from them.The identification of maize haploid seeds has been realized.This system has laid a foundation for the development and improvement of the follow-up researches.