show_detections

giant.ufo.visualizer:

giant.ufo.visualizer.show_detections(ufos, camera, save_frames=False, interactive=True, frame_output='./{}.png', log_scale=False)[source]

This function plots possible UFO detections for each turned on image in camera over top of the image.

For each individual image, the ufo results contained in the ufos dataframe are plotted as a scatter plot colored by their quality code. The image itself is displayed as normal, or possibly using a “log” scale to bring out dimmer features in the image.

There are 3 different options for displaying these figures. The first, if interactive is True shows the images/ufos in a single window where you can navigate from frame to frame using the left/right arrow keys. This is the recommended way to view the results.

The second option for displaying the figures is to save them directly to a file if save_frames is True. The file the figures are saved to is controlled by the frame_output input which should be a string giving the path to save the files to as well as a format specifier {} to replace with the image file.

The final, not-recommended option, is to show all of the figures at once by setting both save_frames and interactive to False. This will make a single figure window for each image and will show them all simultaneously. This can use a ton of memory and is not recommended if you have more than 10 images you processed.

Parameters:

  • ufos (DataFrame) – A dataframe of the UFOs to show from Detector.detection_data_frame.

  • camera (Camera) – The Camera containing the images that the detections were extracted from

  • save_frames (bool) – A boolean flag specifying whether to save the figures to disk instead of displaying them

  • interactive (bool) – A boolean flag specifying whether to interactively walk through the images

  • frame_output (str) – A string with format specifier describing where to save the figures

  • log_scale (bool) – A boolean flag specifying to use a logarithmic scale to display the images, allowing dimmer things to stand out more