The need: We work to work a cadastre using the image of Google Earth in a georeferenced format that is light.
The problem: The ortho that lowers Stitchmaps low in jpg format, the georeference that it brings is not supported by Microstation.
The solution: Download the image with Stitchmaps, synchronize Google Earth with Microstation to import the georeferenced capture and warp against each other.
We are interested in the ecw because it does not occupy an additional file of georeference and where an HMR or Tiff of 200 MB could weigh just 12 MB without losing much in quality. We have Stitchmaps and Microstation PowerMap V8i, for what we have we will do with this although with other programs could be done with fewer steps.
Let's see how it is done:
1. Image download.
This we have done with Stitchmaps, as already explained before. With the exception, that we have drawn a rectangle in Google Earth, so that it is inserted in the capture of the images.
In Google Earth this is done with Add> Polygon, and in style we choose contour with a line thickness of white 1.4. This we will do so, because Microstation can not import a kml file in these versions, except with FME from Bentley Map. But the version Powermap does not bring this functionality, so to create the rectangle we will have to do it drawing on the image.
2. Create a georeferenced dgn.
This is created by doing File> New, and we chose a Seed3D seed. Image import from Google Earth does not work on an 2D file.
Then we must add georeference to the file, that is done with: Tools> Geographics> Select Coordinate System
In the panel we choose From Library, and as this time we are interested in UTM zone 16 North, then we choose:
Library> Projected> World (UTM)> WGS84> UTM84-16N
If this is the system we use the most, we can right click and add to favorites, to be able to access more easily. We make OK and our file is now georeferenced.
3. Capture Google Earth image
To synchronize Microstation with Google Earth we do Tools> Geographics> Follow Google Earth View. This way, our view is reflecting just what is in Google Earth. It is advisable to have oriented the north there and an acceptable approach.
To import the image we do Tools> Geographics> Capture Google Earth Image, we click on the screen and then complete deployment. What we have there is not an image, but a digital terrain model, with the image as property of footwear.
To see the image, we execute the rendering. In order not to complicate where the render buttons are, I will execute it using the text command. Utilities> Key in> render all smooth. See that there is the box that interests us. That image, despite its poor resolution, is georeferenced.
4. Georeferencing the image
For this, first, we will make points in the corners of the georeferenced image. This is done with the dots command, we will do them in green color, with a representative thickness and with a suitable approach so that the corner of the rectangle is seen. If we lose the image, we execute the render command again, and we do not worry about being so exact, we remember that the accuracy of Google Earth is worse than what we could lose here.
Once the points are made, insert the jpg image that we have downloaded with Stitchmaps: File> Raster Manager, then in the panel we choose File> attach> Raster. Do not forget to leave the option active Place Interactively, because we will enter it manually.
We place it inside the box of the gray image, so that we can stretch it from there.
Similarly, we make points to the corners of the rectangle that is in the color image. These will be done in red to notice the difference.
Finally we should have something like this:
To stretch the image, from the Raster Manager panel, we right click on the image, we choose Warp, with the method Affine of more than 3 points. Then select each corner, indicating the point of origin (red) to the destination point (green) and when they are already the four we make the right mouse button.
5. Convert image from jpg to ecw
Okay, now our jpg image is georeferenced. To save it in another format, select, right-click and choose save as. We can choose from many formats, including the precious Ecw that they did not have the Microstation versions.
And finally we have what we needed, a size 24 MB raster, with a box of our interest of 1225 meters per side, ready to work.