| With improving in earth-observing, satellites obtain more and more data as a result that more and more information can be gained from high resolution remote sensing images. By detecting the cloud over the target, satellites can avoid the shelter from cloud and save energy. This paper concentrates on the research of cloud time window and surficial object time window calculation, improves the operational efficiency of satellites, and has impact on the research of satellite task autonomous planning.This paper first describes the development of cloud heights detection and cloud detection, and then, introduces the key technologies in calculation for time window of cloud shelter, including cloud detection, features detection and matching, the transformation relation of images and the retrieve of fundamental matrix, and tests their performance.To satisfy the real-time calculation for time window detection, a method of high-speed cloud-top heights detection is proposed. On the basis of the difference in matching relationship between cloud and ground with image couple, projecting deviation of cloud to ground is computed and then cloud-top height is calculated by geometric relation. The results of experiments and MISR images suggest that the bias of calculation is less than a pixel-resolution. Error analysis shows the trend of relative error along with different imaging angles.Finally, this paper demonstrates the complete cloud time window calculation algorithm and example. And then, this paper introduce the structure of DEM data, the mathematical model of surficial object time window detection as well as the algorithm and example. At last, the software based on the proposed method is demonstrated. |
