Block adjustment of zy-3 multi-temporal images without ground control points

The ZiYuan-3 (ZY-3) satellite is China's first civilian high-resolution stereo mapping satellite. The revisit cycle of the ZY-3 satellite is only five days, which means that ZY-3 multi-temporal images are easily available. In this study, the block adjustment of 12 multi-temporal datasets of ZY-3 three-line camera images without ground control points (GCPs) is tested and reported. The bias-compensated rational function model (RFM) is used as the block adjustment model. The RFM-based free network adjustment is introduced in detail. The experimental results showed that the block adjustment accuracy achieved with ZY-3 mono-temporal images reached an average of only 18.966 m in planimetry and 7.709 m in height. When these multi-temporal images were used together in a block adjustment, the redundant observations of ground points were undoubtedly increased. Accordingly, these redundant observations could increase the randomness of residual errors of ground points. After the block adjustment without GCPs, some of the residual errors of ground points were effectively self-eliminated. Then, the block adjustment accuracy of ZY-3 multi-temporal images reached 7.468 m in planimetry and 2.064 m in height.