基于GVG采样系统的农作物估产可信度分析——以四川盆地为例

Reliability Analysis of the Accuracy of Crop Yield Estimation Based on GVG Sampling System

以四川盆地的农作物估产为例,对GVG(GIS&VIDEO&GPS)采样系统的采样结果进行可信度分析。在试验区随机选取具有代表性的采样线,利用GVG采样系统提取样区内农作物种植比例。然后结合野外实地调绘结果对该地区进行遥感影像解译,在解译影像上对采样线不同缓冲区范围内的农作物面积进行提取,计算出农作物的种植比例。把遥感解译影像缓冲区分析结果为检验数据,采用误差最小可信度最高的原则对GVG采样系统采样结果的精度进行可信度分析。并基于可信度分析初步确定GVG采样系统的沿线采样最佳范围。

GVG sampling system is a filed sampling system that integrates the GPS, GIS and video technologies. It mainly photos the crop on the side of the sampling line taking the method of linear sampling, which could reflect how the crop grows. Then we can calculate the proportion of every kind of crop from the photos using the GIS model. To analyze the reliability of the accuracy of GVG sampling system, the article compares the result of GVG sampling system with the one from image interpretation, taking crop yield estimation of Sichuan basin as an example. The process of getting proportion of crop area includes two steps. Firstly, interpreting the remotely sensed image of research region and taking the results of field investigation into account, then finding the certain location of sampling line on the image. Secondly, making buffer analysis to the line as well as calculating the crop area within the buffers. And fitted analysis is mainly on the principle that the smaller the fitted error is the more reliable the result is. In addition, the best range of the sampling line is also suggested, within which the results of GVG sampling system are reliable.