Research On The Detection Method Of Egg Embryo Activity Based On Near-infrared Light

Egg embryo is the carrier of most avian influenza vaccines and some human vaccines.Egg embryos will die during the culture process.Dead egg embryos will become a source of bacteria and fungi,causing cross-infection of egg embryos and polluting the entire production environment.Through the investigation,sorting,comparison and analysis of high-throughput detection technology,this subject regards the detection of weak biological signals and the classification of egg embryo activity as the key problems to be solved in this research,and solves the poor quality of collected signals based on fast digital phase lock technology.During the high-throughput acquisition process,the egg embryo signals are interfering with each other,and the activity of the egg embryo is classified based on the entropy method and artificial intelligence.The research of this technology can improve the detection efficiency and detection accuracy of egg embryo in the egg embryo vaccine industry.First of all,aiming at the problem of mutual interference between egg embryo signals in the process of egg embryo weak signal detection and high-throughput detection,the light source modulation and digital phase lock technology are studied,and the egg embryo pulse wave signal signal-to-noise ratio of fast phase lock technology is carried out.analyse and compare.In the detection process,the original signal is superimposed with a noise level of 1 times the noise.The signal-to-noise ratio of the phase-locked signal result and the original signal plus noise are 28.69 dB and18.07 dB,respectively.Aiming at the problem of mutual interference between egg embryo signals,a crosstalk signal with a crosstalk level of 1 time is added to channel 1,a noise and interference signal with a noise level of 1 time,and the signal quality after the phase lock technology is basically the same as the original signal.The study found that the difference in modulation frequency between adjacent channels affects signal detection.The higher the modulation frequency,the better the signal quality of the egg embryo pulse wave.The key technology mentioned above is verified by simulation that the digital phase-locking technology can be used in the egg embryo signal acquisition process,and the simulation experiment provides the basic conditions for the establishment of the acquisition platform.Build an anti-interference platform based on digital fast phaselocking technology to verify whether the anti-interference technology is effective in the egg embryo detection process.The egg embryo activity detection and acquisition platform is designed according to the single-channel anti-interference technology platform,and the signal quality is compared when multiple channels are simultaneously detected.The acquired pulse wave signal of egg embryo can support the research of detection algorithm.Finally,the egg embryo activity detection algorithm is studied on the pulse wave signal obtained by the egg embryo acquisition platform.In order to meet the needs of rapid and high-accuracy detection,a rapid classification study of egg embryo activity is done based on entropy value.Among them,the accuracy of the two classifications of non-rejected embryos and rejected embryos is 99.49%.Based on the study of egg embryo classification based on convolutional neural network,the classification accuracy rates of live embryos,weak embryos and dead embryos are 99.74%,84.39%,and 93.73%,respectively.The research results show that the pulse wave of egg embryos can be collected based on the digital phase-locking technology,the frequency-division modulation method can solve the interference problem between channels,and the egg embryo detection algorithm can realize the accurate classification of egg embryos.