Collection And Analysis Of Field Sampling Points For Land Cover Monitoring Based On Remote Sensing

Field sampling points are an important source for the training sample and accuracy assessment sample in the process of the land cover mapping.The quality of field sampling points can contribute to the land cover products with high degreen of accuracy. It can provide scientific basis for the monitoring of the land cover dynamic change. The technologies of GPS, GIS, mass storage and wireless network are integrated in mobile sampling. Many researchers focuses on mobile sampling due to the efficiency and high-precision of mobile sampling. To develop the sampling system of field land coversampling points, this paper thoroughly studies the theory and means of contactless positioning and applies compass and laser ranging technique for land cover field surveys. Furthermore, this study utilizes the terrain data and pre-existing land cover dataset to assess the representativeness of the field sampling points based on statistical approaches and spatial analysis methods.According to the field survey of land cover and the analysis of land cover sampling points, this study achieves several following conclusions:(1)This paper developed a set of software, which integrated GPS compass and laser ranging technique to measure the coordinate of the stand-off objects based on the theory of intersection survey and traverse survey. The result shows that the operator can use this software to measure the coordinate of the the stand-off objectsin the range of 200 meters, without contacted with the targets. Measurement accuracy of the contactless positioning is under 30 m, in range of one pixel. That is to say, when moderate or low resolution remote sensing images were used to map land cover products, the proposed software can acquire high-accuracy geospatial information, which satisfies the demand of field sampling. (2)In field survey, this paper utilized the proposed software to acquire the field sampling points of land cover based on high-resolution images, pre-existing land cover datasets and professional knowledge in remote sensing. The result showed that it can efficiently and accurately collect the field sampling points to apply this software for land cover mapping. (3)This paper calculated the aggregation index and the coefficient based on Delaunay triangulation and Voronoi diagram to assess the representativeness of the field land cover sampling points. Assess results show that the representativeness of the dataset of filed land cover sampling points,which was collected by the proposed software, have spatial representativeness. (4) This paper built the rules between he terrain data and land cover based on DEM, slope and pre-existing land cover dataset to guide the field survey. Furthermore, according to the above mentioned rules, this paper used the software based on remote sensing and correlation coefficient to assess the representativeness of the field sampling points.