Python3 version of ClaRAN
All information about requirements, setup, tutorial, etc. of ClaRAN code can be found on the link above.
After generating cutouts following approach outlined in ClaRAN's paper.
Run: cd tools and %run /users/cmofokeng/rgz_rcnn/tools/demo_gmrt.py --radio /idia/users/cmofokeng/data/gmrt/split_fits/test_1deg/gmrt_en1w610_clipped_21.fits --ir /idia/users/cmofokeng/data/gmrt/split_fits/test_1deg/gmrt_en1w610_clipped_21_infrared.png --catalog /idia/users/cmofokeng/data/gmrt/en1w610-5sg9-clean-offset.vot
catalog argument was used in order to overlay positional information about the centre of the cutout. The script above returns a csv file of detections from ClaRAN.
Switch to sc_pipeline branch for the pipeline code.
From the detection and classification phase, ClaRAN ouputs the csv files of the detections which were grouped into a catalog.
The bounding boxes of the detected sources are used to find the positions of the radio source, furthermore the estimated source positions are then used to cross-identify IR sources.
Run: python3 source_positions.py to characterize sources.
PS: the code currently uses static files, the next version will improve upon this.