Study On Dynamic Monitoring Of Cotton Spider Mites Based On Remote Sensing Of UAV

Cotton is the main cash crop in Xinjiang.In recent years,with the constant changes in the climate environment,the cotton field ecological environment has become more conducive to the occurrence of pests and diseases,and cotton spider mites as one of the main pests on cotton,is extremely harmful to cotton production.How to quickly and accurately monitor cotton spider mites and timely grasp the occurrence dynamics of field damage is of great significance for reducing the loss of cotton production.Based on this,the paper uses the advantages of UAV remote sensing in spatial resolution and spectral resolution to obtain real-time and dynamic information on the occurrence and development of cotton field damage,combined with the changing characteristics and meteorological elements of the damaged cotton field,from remote sensing identification of spider mites,regional dynamic monitoring,impact factor analysis in three aspects to carry out cotton hazard monitoring research.The main research contents and research results are as follows:(1)The research on the identification of cotton field damage in multi-spectral images of drones provides a basis for monitoring the damage.By looking for the spectral index with significant relationship with the spider mites as the primary characteristic factor,the corresponding feature values were extracted from the ground damage data and the image data of the affected cotton field.At the same time,the Akaike information criterion was used as the basis for feature selection,and the best construction was obtained.Model features,a logistic regression model for spider mite identification was established.Studies have shown that among the total spectral indices analyzed,TVI,DVI and RDVI are the best characterization factors for identifying spider mites.The logistic regression model based on the three factors has a classification accuracy of 95% and an F1 value of 95.1%.It can better realize the identification of cotton spider mites and lay a theoretical foundation for the research on the monitoring of damage.(2)The method for monitoring cotton spider mites in multi-temporal remote sensing images was developed.Based on the multi-temporal image data acquired by the UAV,the spider mite identification model obtained in this paper is used to extract the damage information of each period of image.On the basis of this,the image information statistical method and the change detection method are used to analyze the occurrence and development of the damage time and space.The process,and using the data interpolation method to calculate the area of the damage occurrence in the time series,establish an exponential curve monitoring model of the damage over time,and realize the dynamic monitoring of the area of the damage.(3)Conduct quantitative research on the impact of climate environment on the occurrence and development of spider mites.The quantitative relationship model between meteorological factors(average temperature,maximum temperature,minimum temperature,relative humidity,rainfall,temperature and humidity coefficient,temperature and rainfall coefficient,accumulated temperature,wind power and wind direction)and spider mites was constructed.The results showed that the wind,warm rain coefficient,temperature and humidity coefficient and accumulated temperature four meteorological factors affected the dynamics of the spider mites damage,and the wind direction affected the change trend of the spider mites.Based on UAV remote sensing data,the cotton leaf mites in cotton fields were identified and monitored rapidly in a certain area.Combined with various meteorological data,the occurrence and change process of cotton mite were analyzed.The research results can provide methodological reference for accurate monitoring and trend prediction of cotton mite damage,and provide basis for unified control and early prevention.