多时刻海色遥感数据融合及其可视化

Data Fusion and Visualization of Multi-time Ocean Color Remote Sensing Data

为解决MODIS、SeaWiFs等海色传感器数据容易出现空洞的问题,提出一种利用多时刻数据结合多种拟合方式自适应修复的算法。针对前后时刻中对应位置含有有效数据的空缺位置,设置阈值A,有效数据量大于A的空缺点,采用最小二乘法拟合的方式进行空缺值估计,否则采用线性插值法进行估计;对于所有时刻均为无有效值的空缺位置,采用同一时刻相邻位置的有效数据进行双线性插值法进行第二次修复。结果表明:在阈值选择合适的情况下,该算法可视化效果平滑,数据分析结果误差相对较小,并能实现全球数据的较好修复,提高数据有效覆盖度。

In order to solve the problem that the ocean sensor data such as MODIS and SeaWiFs are easy to emptied,an algorithm combining multi-time data with multiple fitting methods is proposed. The vacancy value is estimated by the least squares fitting method.To set a threshold named A according to the date of different time in corresponding position.When the effective value is larger than A that the vacancy value is estimated by the method of least squares fitting,when the effective value is smaller than A that the linear interpolation method is used to estimate the vacancy value.There is no valid value the bilinear interpolation method is used for the second repair with the valid data of the adjacent position.The experimental results show that the algorithm has a smooth visual effect when the threshold selection is appropriate, and the error of the data analysis results is relatively small,then the global data can be well restored and the effectiveness of data coverage will be improved.