When I upload TIFF images to QGIS, all of these are placed on top of each other. The files are "telling" me they dont have the right projection, but I have tried many (including the same as my World map is in) but its still messed up. I have managed to change the placement of the images, but never to the correct location. And they are still placed on top of each other. My goal is to make NDVI analyses from these. How do I manage to place them at their correct locations? I have taken the images with a DJI Phantom IV RTK drone, and all images have GPS coordinates in their files.
-
1Your images are not georeferenced. It means that QGIS does not know where to place them in the coordinate system and because of that the default placement is used and the top-left corner of the image is placed into (0,0). QGIS and GDAL in behind it do not use the coordinates which are stored as EXIF data into digital images. That data alone is not enough for georeferencing because pixel size, orientation etc. is missing. You must georeference your drone images with some other utility first.– user30184Commented Oct 28, 2020 at 7:34
Add a comment
|
1 Answer
To expand on @user30184 . The EXIF data stored in the image is the location of the GPS/camera at the time of image capture. There are several QGIS plugins that will allow you to map the locations of those points but the points will only be the locations of the camera at capture. These images will still not be georeferenced. These plugins typically allow you to click on the point in QGIS and see the image in an image viewer, but the top edge of the image will be relative to the top of the camera at capture and will not be rotated base on a compass bearing. The image will not be skewed based on the pitch, roll, and yawl of the camera at the time of capture.
Consider using WebODM (multi-platform FOSS4G), Agisoft Metashape (multi-platform), Pix4D (cloud), or DroneDeploy (cloud) - all with free trials - to generate an orthomosic image. These structure-from-motion (SFM) application will look at the EXIF data, and automate the stitching of the images to a single georeference orthomosaic. Some SFM packages will even output an NDVI image for you! Be warned, these are complicated software packages and require some experience and knowledge (despite what the drone companies will have you believe.)
These SFM software packages look for items in adjacent images and use those objects to paste the individual images together. If your images cover a large area of row crops, or forested lands - and contain no roads, buildings, water features, etc - the SFM software will not be able to stitch them together.
