Research On The Method Of Rice Remote Sensing Identification And Estimation Of Area Based On Spectral Time-series Fitting In Southern China

Food security is an important guarantee for the stable development of our country,and the grain acreage is the basis of food security,so the estimation of grain acreage is important.Remote sensing technology is an important means to estimate crop grain area at present.The classification accuracy is affected by cloud and mist,which is not avoided.To solve this problem,this study presented a method for recognizing rice based on GF-1 time-series image.With long time-series of GF-1 images,three indexes about middle-season rice and late-season rice,namely near infrared band reflectance(NIR),red(R)band reflectance and the normalized difference vegetation index(NDVI)characteristics are extracted.Fitting spectrum and vegetation index time-series characteristic curve,and analyzing for image discrete value of near infrared band,red light band and NDVI fell on both sides of the proportion of the sensitive area of the fitting NIR,R and NDVI time-series curve of middle-season rice and late-season rice.The area can also be seen as rice crop identification features of target area,and only those reaching a certain proportion can be identified as certain type of rice.In this case,three kinds of comprehensive situation voted to decide final classification results.The means of samples are used to fit the curve for each image.The outliers are eliminated with the ground samples in advance.Statistical analysis of ground samples defined target characteristics.In this paper,the statistics field investigation in the quadrat rice planting area,part of the samples as the revised samples,the remaining samples for validation area estimation precision,establishthe linear relationship between WFV classification rice area and rice area of the samples,the rice area is calculated by linear regression model correction parameters,using linear correction model to estimate rice area inside the validation samples,compared with real rice areainside validation samples,verify the accuracy and the last for the whole study area GF-1 WFV rice classification statistic results are correct,estimate the rice planting area.The result indicated that : this method can identify middle-season rice and late-seasonrice accurately in cloud and mist area,middle-season rice and late-season user accuracy can reach 95.97%,95.95%,the overall accuracy is 95.76%,the kappa coefficient is 0.9335;and the experimental results are ideal,Showed the effectiveness of NIR,R and NDVI time series curve fitting,and the rationality of target locale of the fitting curve.This methodsolved the problem of cloudy foggy in the recognition of rice,which laid a foundation for rice planting area estimation.