变化信息监测的时域IHS变换

MULTI-TEMPORAL HIS TRANSFORM FOR CHANGE INFORMATION DETECTION

在IHS变换过程中 ,对I分量图像与高分辨率全色波段图像采用线性回归辐射配准 ,减少了不同传感器、不同时相影像间光谱噪音 ;采用相关系数法叠加方式 ,生成一新明度分量进行反变换 ,增强了变化信息 ,保留了多光谱信息与纹理信息 ,有助于自动分类。

In order to applying IHS transform to change detection, first use linear aggression to perform I component and high spatial resolution PAN radiometric rectification to minimize noise and variation caused by the difference of sensor type and imaging time. Then use coefficent overlay method to create a new I component for inverse transform. This method improves the fusion result and enhance the change information for classification.