Montoring Farming System In A Larger Area Using Remote Sensing Technology And Open Service System

China’s agriculture has entered a new level of development,and more and more new technologies were applied to the practices of agricultural production.Remote sensing technology has become one of the necessary means of the current regional agriculture research with its quick efficiency,large-scale and multi-spectrum information obtained.In order to better obtain planting information of agricultural crops,standardizing the process of analysising crops with remote sensing technology,further enhance the operational capability of remote sensing and study farming system deeply,we builded the platform of crop remote sensing open service system.The paper used GF-WFV remote sensing image data as data source,used Image Processing Machine(IPM)to complete the control point matching,adjustment and image optimization,batched geometric correction and orthorectification of images to correct image geometric deviations and orthophotos deviation;As a result of crop growth environment and satellite image sensor radiance differences,phased radiation normalization was arranged,based on pseudo-invariant feature point method for the initial relative radiation normalization correction,then mapped image by NDVI grading according to the principle of natural breakpoints between reference images and warp images to be corrected and generating new image to achieve normalized radiation correction;Taked into account the growth characteristics of crops,the feature images with growth index are generated to complete the conversion of image data.For the large amount of crops image data during the multiple growth period,we can use the time grid as a reference to construct the spatial storage mechanism with the help of raster model,achieving fast transfer and calculation by distributed service strategy;Drawing multi-temporal growth characteristics curves using seasonal characteristics of crops.The paper researched that how to distinguish crop by interactive decision,which samples were projected onto the growth feature images grading to obtain the threshold.In this way,crop planting information analysis and application will be used to provide basic crop planting information for the research of cropping system.Taking the distinguishing of bulk crops in Jiangsu as an example,this paper tests the applicability and practicability of the wide-range monitoring methods.The result shows that the application of the method is efficient and reliable,and can be used for analysising crop planting information based on remote sensing.The paper solved the following three problems:(1)Efficient storage and processing of image data.The method for massive image data processing is more efficient than traditional methods,also can save a lot of time.The paper solved the problem of storing and tranferring a large amount of raster and vector data.The distributed system service strategy maked full use of the computing resources of each server,and also provides a convenient channel for the later function expansion of the system.(2)Model threshold by interactive decision-making.The segmentation gradings of the growth index feature image value domian combined with some sample data were analysised,and the user interaction was used to determine the crop identification threshold quickly based on remote sensing.The method of threshold determined by deliverring sample effectively reduced the subjectivity brought by human-computer interaction,increased the reliability of the model and realized the soft segmentation of the image.(3)Open remote sensing monitoring service.Applying wide range monitoring based on remote sensing method to build farming systems services system,immobilizing of monitoring processes,placing the remote sensing image processing in the background,opening the model parts for agricultural workers.System can make full use of practitioners’ agricultural experience,and get key parameters of model via supervision data interacts with workers.Therefore,system can achieve rapid and efficient extraction of regional crop cultivation information,and provide necessary tools and conditions for the monitoring and research of regional farming systems.