For some graduate work, I am collecting data and using it to train AI object detection models (such as Faster R-CNNs, EfficientDet series, and others that can be found on the Tensorflow 2 Detection Model Zoo). This process involves collecting images with objects of interest and labeling them with an open-sourced image labeling tool (Labelimg). After training, these deep learning models can be exported and used to locate objects in images by whomever has access to the models.
My question is this: Is there a good software platform that would allow for the exported models to locate objects on image orthomosaics, such as those taken by aerial imagery? It is possible to use python and
matplotlib to detect objects on individual images, but this method doesn't work well for larger orthomosaic images.
The reason I need to get GPS coordinates of the resulting bounding boxes in order to export them for further use. I was thinking that there was a way to upload the model to ArcGIS Pro or QGIS to do this, but wasn't sure if this would be possible or not. Is there a good way to do this?