Hello all! Need some advice here. I noticed since yesterday, half of the satellite imagery in Chaozhou, Guangdong has been shifted 7m horizontally and 5m vertically. Without presence of GPS points, I can’t determine which half of the imagery is actually on the right position. For that, I can only think of using one of the following solution to proceed editing:
Use Aerial Shifter script to move new imagery to match existing road segments
Move existing road segments to match new imagery (this involved more than 10k number of road segments :shock: )
Leave existing road segment as it is. For new segment from this point onward, all created based on new imagery.
Personally I prefer option 1. But result of the segment position will be 50% correct or 100% wrong… Please advice if you guys have other solution. Thanks!
I couldn’t see any shifted road. They line up perfectly with the aerial view and GPS trace. Could you give us a permalink so we can double check it please?
Hello Leo. I have identified the area of shifted imagery. It only covers considerably a small area. The shifting starts from here, and the displacement will decrease and eventually joint up with surrounding imagery as panning the map toward north . I have put up two temporary boundary markers to remind myself to work with Aerial Shifter Script on: boundary1 & boundary2
Please advice. Thanks!
Interestingly I couldn’t see any shifting. They are perfectly matched each others. However, I couldn’t see any GPS traces for these area too so I’m not sure what’s going on. Could you try to turn off all the scripts and refresh the screen and see any differences.
I assume the existing segments are correct and should not be changed. If the aerial view has been shifted, we never follow the av. Always check with the GPS trace. If the av is shifted on your screen, please use option 1. Don’t do option 2 or 3.
I noticed since yesterday, half of the satellite imagery in Chaozhou, Guangdong has been shifted 7m horizontally and 5m vertically. Without presence of GPS points, I can’t determine which half of the imagery is actually on the right position. For that, I can only think of using one of the following solution to proceed editing: