针对“天绘一号”01、02星高分辨率相机品字形CCD机械结构的特点, 分析了引起高分辨率片间CCD图像错位的影响因素, 利用实验数据定量描述了片间CCD图像错位现象。提出了基于特征的图像重叠区域同名点自动提取和优化方法, 针对错位帧数据建立了错位拟合函数修正模型。预处理试验表明本文所述方法有效解决了高分辨率相机的图像错位问题, 实现了CCD图像的无缝拼接。

Aim to the mechanical structure of Mapping Satellite-1 high resolution camera, the impact of factors caused the dislocation of the CCD image in high resolution slice is systematic analyzed, experimental data is used to quantitatively describe the chip CCD image dislocation phenomenon. Automatic extraction and optimization method is proposed for same name point in the image overlap region based on the characteristics, the dislocation fitting function correction model is established for the dislocation frame data. The preprocessing production tests show that the method is an effective solution to the high-resolution camera image dislocation problem and CCD image seamless mosaic.