HJ-1B卫星红外扫描图像的几何定位和精校正方法研究

Research on geometric correction method for infrared scanning camera of HJ-1B small satellite

根据环境与灾害监测预报小卫星1B星(HJ-1B)携带的红外扫描相机的成像几何特性,提出了一种基于成像过程模拟的几何定位和精纠正方法。该方法首先根据共线方程对HJ-1B红外相机的图像初定位,再基于一幅更高分辨率的已校正的遥感图像,按照红外相机的成像几何模拟生成若干图像块,与待纠正的红外相机图像匹配,用最小二乘修正偏移量,最后根据修正量去校正有误差的星历参数,从而获得整幅图像的精确定位信息。经过模拟数据和MODIS数据的检验,所提出的算法是可行的,定位精度有明显的提高。

A geo-location algorithm for the infrared scanning camera in the HJ-1 B small satellite is proposed in this paper based on pixel navigation and image matching. The algorithm is composed of three steps. The first step is the pixel navigation based on the principles of photograrnmetry using the parameters of the satellite ephemeris and attitude information. The second step is to simulate the HJ-1B satellite image with an existing remote sensing image, which is accurately georectifled and of higher spatial resolution than the HJ-1B image. The HJ-1B image is then matched to the simulated image using the least-square matching method to find the accurate location of each HJ-1B image pixel. After geo-location, the image can be geo-rectified according to the location of each pixel. In the preliminary test with the simulated image, the geometric correction method is proved feasible in principle as well as in practice.