I'm working with Mask R-CNN to locate fires.

I have the following image. I have info about the geographic center coordinates of the image, and the orientation of the image in degrees let's suppose this image is oriented to north. I have also the center coordinates of the masks. How can I convert the center coordinates of the mask in pixels to geographic coordinates.

example image of fires


If I understand correctly, the pixel coordinates of the small images are referring to the main image. And I assume you know Python.

I would start by rotating the main image to be perpendicular to axis and then apply the same rotation to the other images. Then, using the coordinates of the upper left corner and resolution (pixel size), set up the geotransformation of the image using rasterio and Affine, and then you will be able to compute geographic coordinates of any pixel in the image. (You would check for the centers of the small images.)


There are many easily identifiable points, so I suggest to use georeferencing tools. It is available in qgis. https://www.qgistutorials.com/en/docs/advanced_georeferencing.html

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.