i am playing around with some CNNs for semantic segmentation. The job is to detect all pixels that belong to buildings from aerial footage. This is a binary semantic segmentation, so i am only interested in "building" and "not building". I have got huge trainingsets for this purpose and i am wondering if it's possible to turn these trainingssets to sets for object detection. So a program, that takes an image with outlines of all the buildings and just creates the bounding boxes of those buildings. That should be quite easy i think...

Thank you in advance for your help! Knust

Your Answer

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

Browse other questions tagged or ask your own question.