| With the continuous development and reform of remote sensing technology,remote sensing technology in the field of crops has also emerged.Hunan has a long history of rice cultivation.In the rice-growing counties in Hunan,Xiangtan is a rice-growing county in Hunan.The area is dominated by double-season rice and mid-season rice cultivation.The cultivation method is mainly based on throwing rice fields..The artificial field survey of rice faces the problems of large manual workload,low accuracy,high cost and slow speed.Compared with remote sensing monitoring,it has the advantages of fast and convenient,high accuracy.Rice production has a large proportion in China’s grain output.If you can use real-time monitoring of rice by using remote sensing,it will greatly help to reduce manpower,material resources and financial resources,and improve the agricultural economic development and food-related policies.Has a major impact.Rice planting information monitoring includes two important aspects of crop identification and planting area extraction,growth monitoring and yield forecasting.Crop identification is based on the principle of different spectral characteristics of different crops,using the difference in reflectance in the near-infrared band and multi-temporal remote sensing to effectively identify crops.Remote sensing rice growth monitoring utilizes the near-infrared and red-band remote sensing information,and then the crop normalized vegetation index(NDVI)is positively correlated with the crop leaf area index(LAI)and biomass,and the main selection study here.The object is the normalized vegetation index(NDVI)curve of the crop,and the normalized vegetation index(NDVI)curve of the crop is obtained from the remote sensing image,and then the leaf area index(LAI)of the crop is calculated in reverse,thereby realizing the crop.The growth of the situation.The extraction of rice planting area refers to the use of remote sensing technology to extract relevant information with the support of GIS,so as to obtain the planting area of rice.Crop yield estimation is estimated by remote sensing.Different pop crops have different pop reflection characteristics.Remote sensing estimation uses this principle and uses remote sensing technology to monitor and forecast the yield of the crops under study.Ways.Based on the Landsat 8 remote sensing image as the main data,this paper extracts the planting information of the late rice in Xiangtan area in 2015 and 2016 respectively.The main research contents are as follows:Remote sensing image preprocessing in the study area: The DN is converted into a relative value related to the physical factors such as the reflectivity of the earth’s surface and the surface temperature by a mathematical method through radiometric calibration.Then,through atmospheric correction,the effect of removing part of the atmosphere is removed,and a more realistic reflection rate of the ground object is obtained.Separability analysis of remote sensing images: Using Landsat 8 remote sensing image data as the main data source,selecting feature points as training samples,and verifying the accuracy of the training samples we extracted,to ensure the accuracy,and then using Xiangtan phenological data,Xiangtan Some indicators such as regional topographical features and vegetation coverage(NDVI index)in Xiangtan area are used as auxiliary data to extract the six key features of waters,cultivated land,construction land,paddy fields,grassland and woodland that we need in our research.Classes perform separability analysis and spectral analysis of key features.We will use the sum of the pixels in the study area obtained by ENVI to calculate the rice planting area in Xiangtan,and compare the calculated data with the statistical data to calculate the planting area of rice late rice in 2015 and 2016.Combined with actual statistics,the overall accuracy in 2015 was 87.6923%,and the Kappa coefficient was 0.7533.In 2016,its overall accuracy reached 85.8407%,and the Kappa coefficient was 0.8273... |
PDF Full Text Request