梯度信息在影像数据正常高处理中的应用

The application of gradient model for normal height of the image processing

针对当前影像数据处理中正常高精度存在冗余的问题,将梯度理论引入区域似大地水准面格网模型中,为正常高的获取提供轻量化、实用化模型。基于似大地水准面模型计算区域高程异常的梯度;基于不同比例尺图幅正常高的精度限差,推算相应图幅内高程异常的梯度限差;最后确定满足梯度限差要求的图幅。将CQG2000的梯度数值应用于1∶25000图幅中,结果表明:全国约99.14%图幅的梯度能够满足高程精度要求,可以以图幅中心点的高程异常值代替整个图幅的高程异常;仅在昆仑山、天山及青海局部等区域共计约0.86%图幅不满足要求,需按照双线性内插等常规方法计算图幅高程异常值。

In view of the redundant problem of normal height precision in image data processing,the gradient theory was introduced into the regional quasigeoid model to provide a lightweight and practical model for the normal height.Firstly,the gradient of the regional height anomaly was calculated based on the quasigeoid model.Secondly,the gradient limit of the height anomaly was estimated for the map sheets of the different resolution based on the normal height limit.Finally,the map sheets that meet the gradient tolerance requirements can be determined.Analyze the applicability for gradient value of CQG2000 in the 1∶25000 map sheet,the results show that the gradient of 99.14%map sheets could meet the tolerance of the height accuracy,thus the height anomaly of the entire map sheet could be represented by the height anomaly of the center point of the corresponding map sheet.The gradient of 0.86%map sheets which located in Kunlun,Tianshan and the partial area of Qinghai had the poor accuracy,therefore,it is necessary to calculate the height abnormal of these map sheets according to the conventional method such as bilinear interpolation and so on.