Development of GPS orbit forecasting system for differential positioning in forest area

与农田或海洋相比，森林地区的GPS定位是困难的。主要原因是地形和树冠切断了GPS信号。因此，需要制定GPS测量计划，通过了解GPS卫星何时会通过特定定位点的树冠间隙，来了解该点何时可以进行GPS定位。本研究旨在结合点照片和GPS轨道信息，开发用于林区差分定位的GPS轨道预报系统。研究领域设在北海道的东京大学森林和千叶的东京大学森林。在森林中的每个定位点，拍摄了一张高分辨率的天空鱼眼数码照片和一个多小时的全球定位系统信号。利用数字图像分析技术对数码鱼眼照片进行分析，对照片中的天空、树叶、树枝、地面等要素进行分类。然后这些元素被转换成“ 关于我们 开放度”，其表示GPS信号的穿透程度，以生成“开放度”图像。根据历书数据计算出观测卫星的GPS轨道，并将其绘制在“开放度”图像上，以提取每10秒的“开放度”数据。分析每个GPS卫星的信号，以计算每10秒的信号强度。然后以“时间”为关键词，将“开放度”数据和信号强度数据进行组合，并分析了两者之间的关系。结果表明，“开放性”与信号强度基本相关，但不能完全解释信号强度的波动。这是由于透镜不水平、北向定位误差等不良摄影条件造成鱼眼照片与GPS轨道不对应的结果。但在森林中拍摄时，这些误差往往会发生。为此，开发了一个在用户自定义“开放度”值和时间间隔下预测GPS卫星数量的程序。通过该程序，用户可以预测他（或她）什么时候可以做GPS定位在特定的点，如果他（或她）采取鱼眼照片的那个点。这将极大地促进GPS在林区的应用。少

GPS positioning is difficult in forest area compared to agricultural field or ocean. The main reason for this supposed to be cut-off of GPS signal by terrain and tree canopies. So, it is needed to make GPS survey plan, by knowing when GPS satellites will pass through tree canopy gap of particular positioning point, to find out when GPS positioning is possible at that point. This research aims to develop GPS orbit forecasting system for differential positioning in forest area, by combining photograph of the point and GPS orbital information. Research fields were set at Tokyo University Forest in Hokkaido and Tokyo University Forest in Chiba. In each positioning point in the forest, a high-resolution fish-eye digital photograph of the sky and GPS signal of more than one hour were taken. Digital fish-eye photograph was analyzed by digital image analysis technic, to classify the elements of the photograph, such as sky, leaves, branches, and ground. Then those elements were converted into " … More openness", which denotes degree of penetration of GPS signal, to generate "openness" image. GPS orbits of the observed satellites were calculated from almanac data, and they were plotted on the "openness" image to extract every 10 seconds' "openness" data. Each GPS satellites' signal were analyzed to calculate signal strength of every 10 seconds. Then "openness" data and signal strength data were combined by using "time" as the key, and relationship of these two data were analyzed. The result showed correlation of "openness" and signal strength on the whole, but does not explain all of fluctuation of signal strength. This owes to uncorrespondance of fish-eye photo and GPS orbit, which caused by bad photographing condition such as unhorizontalness of lens and north position error. But these error may happen usually when taking photographs in forest. So, the program was developed for forecasting number of GPS satellites under user-defined "openness" value and time interval. By using this program, the user can forecast when he (or she) can do GPS positioning at particular point, if he (or she) takes fish-eye photo of that point. This will help greatly to promote use of GPS in the forest area. Less