Construction Of Index Recognition Oncomelania Snail Habitats Based On Landsat-8 OLI Imagery

In recent years,China has obtained remarkable achievements towards prevention and treatment of schistosomiasis japonica.Meanwhile,the epidemic situation of schistosomiasis has been controlled effectively and the prevalence kept at a low level nowadays.As the sole intermediate host of S.japonicum,Oncomelania hupensis is one of the most important biotic factors in the transmission and epidemic of schistosomiasis.However,it is alarming that areas of Oncomelania snail habitats have been remained high for the regional throughout the year and new or infective Oncomelania snail habitats were often discovered nowadays,which meaning a serious challenge for restraining the rebound and spread of schistosomiasis.Therefore,exploring and eliminating Oncomelania snail habitats is the most persistent and containable measure for schistosomiasis control,which is also the emphasis and difficulty of schistosomiasis prevention nationwide.With the increased availability in spatial and temporal resolutions from the 1960s,remote sensing technology has shown increased probability for the identification,extraction,and monitoring Oncomelania snail habitats.The traditional snail breeding site identification method is through the vegetation index,water body index,soil moisture and other indexes,and there is no index that can directly reflect the distribution of snail breeding area.Based on the Landsat-8 satellite image,this paper carried out a study on the extraction of Oncomelania snail habitats information,and discussed the feasibility of extracting snail habitat based on the spectral features of the image.This topic selects a typical lake-marsh-type epidemic area—Poyang Lake area in Jiangxi Province as the research site,constructs a snail index based on spectral characteristics,and uses ROC-Jordon index method to perform threshold segmentation on the remote sensing images to predict the breeding site of snails,which has the potential to prevent schistosomiasis.Certain guiding role and practical significance.The main research works are as follows:(1)On the basis of the radiation correction of Landsat-8 remote sensing image data,the actual measured data with and without snail points are synthesized to obtain the spectral characteristics of snail features and snail-free features in the study area,and determine the characteristic bands.The study found that in the near-infrared band(NIR),the reflectivity of snail features is high and that of non-snail features.In the blue,green,and red bands,the reflectivity of snail features is lower than that of non-nail features.The single-band threshold method,ratio method and Difference method,based on near infrared,red and green.The blue band constructs the snail characteristic index,selects the characteristic band and uses the ROC-Yoden index method to determine the characteristic index threshold,segment the remote sensing image according to the characteristic threshold,identify the breeding area of the snail,and evaluate the accuracy.The overall accuracy of the identification of the snail characteristic index 130 reaches 90.23%,the Kappa coefficient is 0.806,which is 1.27%higher than the traditional index of vegetation index,the Kappa coefficient is increased by 0.015,the overall accuracy of the soil moisture index is increased by 8.94%,and the Kappa coefficient is increased by 0.179.The results show that the spectral index construction feature index method can obtain better results in the extraction of Oncomelania snails,better than the traditional index.(2)Using the remote sensing image data of feature index,vegetation index,and soil moisture obtained from multi-temporal remote sensing data from 2014 to 2018,determine the threshold according to the ROC-Yoden index method,segment the image,and determine the snail breeding site in different years and different The spatial distribution during the seasons and the changing law of the snail breeding ground.The study found that the spatial distribution of the snail breeding grounds in different seasons is basically the same as the continental beach area around Poyang Lake.Among them,the identification accuracy of the feature index I30 is the highest in the spring,autumn and winter tests,and the highest recognition accuracy is the feature index 113 in the summer tests;and 2014-2018 The spatial scope of the annual snail breeding ground is decreasing year by year.(3)In order to verify the universality of the constructed snail characteristic index,the measured snail data was collected in parts of the Yangtze River Basin in Anhui Province,and compared with the accuracy of the traditional index.The results show that the recognition accuracy of the feature index is higher than that of the vegetation index,and it is universal.