County-level crop planting area remote sensing statistic sampling survey scheme design method

Abstract

The invention discloses a county-level crop planting area remote sensing statistic sampling survey scheme design method. Multi-period up-to-date GF-1 16-meter-width images are adopted as a data source; a decision tree classification method is constructed to perform automatic interpretation, so that the spatial distribution condition of crops in a research area can be obtained; with the spatial distribution data of the crops adopted as a sampling population, on the basis of the arable land area and crop planting scale of the research area, a PPS sampling method is adopted to select a general survey region; various sample unit sizes and various sample unit quantities are designed in the selected general survey region, the quadrat unit sizes, quadrat sample unit quantities, sampling methods and calculation methods are combined randomly, the calculation results of the statistic sampling survey schemes of different combinations are compared with the planting area of target crops which is obtained through the data interpretation of GF-2 0.8-meter-resolution images; and analysis is performed according to the overall sampling accuracy, sampling stability and minimum sample capacity of result data, and a remote sensing statistic sampling survey scheme most suitable for the target crops of the research area can be obtained.

Description

technical field [0001] The invention relates to the technical field of surveying and mapping remote sensing, in particular to a method for designing a county-level crop planting area remote sensing statistical sampling survey plan. Background technique [0002] Traditional statistical sampling has been maturely used in business operations, but in many ways it cannot meet the needs of modern statistical business. First, the traditional statistical sampling business is based on benchmark data, but the benchmark data often comes from large-scale national conditions and national strength surveys such as the agricultural census, and the data lag is strong, which affects the results of statistical sampling surveys; The sampling method is used to select samples. The samples themselves do not have geospatial information attributes. The crop area data obtained by this method also does not have spatial information, and it is impossible to intuitively present the spatial distribution o...

AI Technical Summary

Problems solved by technology

However, at present, many researches on crop area measurement based on the combination of remote sensing and statistical sampling are based on the whole country, provinces and cities. For area information, there are few studies conducted at the county level

On the other hand, the statistical sampling research based on remote sensing is not systematic. Many studies use a single size of quadrat unit size or a single number of quadrats or a single sampling method for sampling research. The sampling extrapolation method also very simple

Moreover, the crop planting area data obtained by sampling inversion also lacks systematic analysis. The analysis is purely based on the overall inversion error, and has not been evaluated in terms of overall inversion accuracy, sampling stability, and sampling efficiency.

Images