The ZiYuan-3 (ZY-3) surveying satellite is the first high-resolution civilian stereo mapping satellite of China, which carries one multispectral sensor with a spatial resolution of 5.8m. Fully considering the characteristic of the ZY-3 multispectral camera that all CCD chips of four bands scan through the same field, a novel method is proposed in this paper to achieve seamless multispectral images and high accuracy band-to-band registration by re-imaging with virtual CCD. In the experiment, the ZY-3 multispectral camera is calibrated with the ZY-3 images over Henan area and Tianjin area, meanwhile the correctness of the calibration results is validated by 30 high accuracy GCPs in the ZY-3 multispectral image over Hebei. Further, re-imaging method is applied to the ZY-3 images over Hebei area and Lanzhou area with the virtual CCD calculated from the calibrated interior orientation. Finally, the band-to-band registration is estimated by comparing image coordinates of the same points in different bands. The validity of the proposed method is proved by experimental result which indicates the band-to-band registration accuracy is better than 0.2 pixels.
In the process of spatial data integration and matching, the inconsistency often occurs between the rivers and contour lines that generate some spatial conflicts just like “river climbing slope”. Aiming to solve the problem by displacing the contours to matching the river network, this paper put forward a new method for every feature point on contour lines to establish specially its displacement model by affine transformation. The method comprehensively considered relative positions of adjacent ridge and valley points, the angle proportion of different feature points for displacing and local shape of contours to be maintained and transformed, so that the calculation of three constrained parameters determine the displacement of points on the contours that is angle, location and scaling was proposed. To test the method, several experiments were taken to analysis. The results showed that the research can effectively complete the displacement of contour lines to match the river network.
