基于遥感图像的不动产地理信息采集方法研究

Research on Real Estate Geographic Information Acquisition Method Basedon Remote Sensing Image

现有的不动产地理信息采集方法难以精准地划分不动产边界,导致不动产地理信息采集结果误差较大,为此,提出基于遥感图像的不动产地理信息采集方法。以角二阶矩参量和相关性为基础,提取遥感图像的特征参量,根据其划分采集范围内各个不动产的边界信息,通过匹配目标不动产像素在遥感图像空间维度网格范围内位置,实现对其地理信息的采集。测试结果表明,设计方法的Kappa系数均值为0.687,图像信息采集精度均在97%以上,不动产地理信息采集结果与实际信息基本一致,未出现缺失情况,因此说明所提出的地理信息采集方法可以降低采集误差。

The existing methods for collecting real estate geographic information are difficult to accurately divide the boundaries of real estate,resulting in significant errors in the results of real estate geographic information collection.Therefore,a real estate geographic information collection method based on remote sensing images is proposed.Based on the angular second-order moment parameters and correlation,the feature parameters of remote sensing images are extracted,and the boundary information of each real estate within the collection range is divided according to them.By matching the position of the target real estate pixel within the spatial dimension grid range of the remote sensing image,the collection of its geographic information is achieved.The test results show that the average Kappa coefficient of the design method is 0.687,and the accuracy of image information collection is above 97%.The results of real estate geographic information collection are basically consistent with the actual information,and there is no missing situation.Therefore,it indicates that the proposed geographic information collection method can reduce the collection error.