GRASSLAND AMELIORATION BASED ON SPATIAL FREQUENCY ANALYSIS
The objectives of this study were to establish criteria for grassland amelioration and to establish a method for design of grassland amelioration using spatial information on the grassland. To obtain criteria for grassland amelioration, the terrain of the grassland was resolved into spatial frequency by discrete Fourier transform (DFT) using a digital elevation model (DEM). The DEM was generated by a GIS map based on terrain information obtained before and after the grassland amelioration. Then the criteria for grassland amelioration were obtained from the relationship between geographical features and spatial frequency by investigating large differences in spatial frequency before and after the grassland amelioration. Determination of the precise location of the grassland surface is essential for obtaining a GIS map based on terrain information. A real-time kinematical global positioning system (RTK-GPS) was therefore used to obtain terrain information and also to determine the tractors precise position. In addition, an inertial measurement unit (IMU) was used to determine swaying motions (yaw, pitch and roll) of the tractor. These data were used to correct the GPS position and to obtain accurate terrain data. It was found that spatial frequencies showing large differences before and after grassland amelioration were in the range of 0.02 to 0.051/m. Moreover, smooth geographical features of grassland were obtained by deleting spatial frequencies of 0.02～0.051/m from the geographical features before the grassland amelioration. The validity of design data was investigated from design data and amelioration topography data obtained by DFT and IDFT. As a result, the coefficient of determination was 0.7 with a root mean squared error of 0.17. The areas in which field amelioration are necessary can be extracted by the frequency analysis method for geographical features, and the terrain after field amelioration can be predicted by only measuring the geographical features before field amelioration. Then, the design method with a spatial filter was developed in this research analyzing grassland amelioration by the skilled engineer. A RTK-GPS and an IMU were installed in a tractor to obtain data on topographical features before and after field amelioration, and the amplitude change under the spatial frequency domain was analyzed. It was found that amplitude at the frequency of 0.04m-1 (wavelength of 25 m) was reduced dramatically by the field amelioration. A spatial filter was made to mainly eliminate frequency of 0.04m-1. By applying the designed filter, the surface of the land before amelioration became relatively flat, similar to the land after amelioration. Moreover, there was a strong correlation between cut and fill depths of actual ameliorated land and the filtered land. The results suggest that the developed system can be applied to a construction plan of grassland.
grassland amelioration RTK-GPS GIS digital elevation model discrete Fourier transform
Yuji NAMBU Atsuro HAYAMI Noboru NOGUCHI
Foundation The Hokkaido Agricultural Modernization Technology Research Center,JAPAN Kubota,Co.Ltd.,JAPAN Graduate School of Agriculture,Hokkaido University,JAPAN
国际会议
北京
英文
1-6
2009-10-14（万方平台首次上网日期，不代表论文的发表时间）