Human-Computer Interactive Fast Edge Matching of High Resolution Orthophotos

In the process of high resolution images orthorectification, an uneven distribution of ground control points in mountainous area results in the phenomenon of ground objects dislocation between orthophotos edges in the overlap area. Followed image mosaic, fusion and interpretation are seriously affected. High-resolution (2.5 m) panchromatic images with a 30% forward overlap of ALOS satellite were acquired over topography complication zones in Beijing mountainous area. After being ortho-rectified in Geomatica PCI 2012, odd or even serial numbered orthophotos on the same satellite track without edge overlapping were mosaiced and then experimental automatic matching tests between mosaic orthophotos and residual orthophotos were carried out in Automatic Registration of ENVI 4.8 to obtain empirical automatic matching parameters and guiding ground control points in the overlap areas. Finally, automatic edge matching operation was conducted by using parameters and guiding GCPs previously acquired. The results showed that mosaic alternate orthophotos in course line or lateral course lines shortened edge matching time effectively. In Automatic Registration module of ENVI 4.8, Search Window Size = 361, Moving Window Size = 19, Area Chip Size = 256 were empirical automatic matching parameters of ALOS panchromatic orthophotos. This method can accomplish high resolution orthophotos edge matching fast and accurately.