Example 8: Lunar/Solar Surface Video Processing

This example demonstrates PIPP processing a jumpy lunar surface video, though the same operation could be performed on a solar surface video.  The purpose of this process is to stabilise the frames of a video that are too jumpy for stacking software to handle.  In my experience both RegiStax 6 and AutoStakkert!2 actually handle these types of jumpy videos exceptionally well, in fact both programs handle the video used for this example without any problems!  However, I think it is it is worth having this technique available for the odd troublesome video that does trip up the stacking software.

Examples of surface videos that can cause problems for stacking software:
  • Surface videos that have very sudden jumps caused by bumping or kicking the telescope mount.
  • Surface videos that are paused during recording.
  • Surface videos made by joining 2 or more videos together.

Obviously it is worth avoiding these situations but this function may be able to rescue something from the video data rather than just deleting it.

This is the source video that will be used for this example.  Note how the video is constantly being blown around, often taking the features that we are trying to image out of the frame.

Thanks go to Steve Curling (Blazar on SGL) for sharing this video.

As usual PIPP is started and the video file is added to the Image Files List.

Next we select the 'Optimise Options For Solar/Lunar Close-up' option to select suitable options for this time of imaging.

After optimising the options for solar/lunar close up imaging the processing options tab has some changes made to it.  The options changed because of the optimise options selection can be identified by their green text labels.

Note that 'Monochrome Conversion' has been enabled but this will only effect colour videos.  The main thing to note however is that 'Frame Stabilisation Mode' has been changed to 'Surface Feature' which causes a set of 'Surface Stabilisation' controls to be shown and the 'Enable Surface Stabilisation' control has been checked.

Switching over to PIPP's Image Preview window we see that the 'Anchor Feature Box' (AFB) is now shown on the image.  The user must position the AFB around an 'alignment feature' in the frame that PIPP will use to stabilise the video.  A good alignment feature should have light and dark areas and should be distinct so that it can be uniquely matched in subsequent frames.

We place the AFB by dragging its control handles (the small squares).  The centre control handle move the whole AFB around without changing its size and the corner control handles allow the AFB to be resized to match the chosen alignment feature.  Note that larger AFBs can take longer to process.

Now that an AFB has been set we shall process the video to see the effect the current options have on the outcome.
To do this we go to the 'Do Processing' tab and press the 'Start Processing' button.

This is the result when the processing has completed.  At this point we check the Summary details in the output area.

Here we see that from 1433 input frames, 44 have been discarded because the AFB could not be found.
Here is the video generated from this processing:

We see that each fame has been shifted to move the anchor feature to the same position as it was in the first frame and that black borders appear in each frame where there is no available data.  This video is not very useful for stacking as the black borders tend to create artefacts in the stacked image.

In order create a video more suited for stacking an 'Area Of Interest' (AOI) should also be used.  The AOI can be used to specify an area that must be present in order for a frame not to be rejected and it can also be used to specify the area to crop the video down to.  In this example we will both reject frames without a complete AOI and also crop the final video to the size of the AOI.  This will mean the generated video will not have any black borders at the expense of rejecting more frames.

We continue by ticking the 'Enable Area Of Interest' control.

Now PIPP's Image Preview Window also shows a blue 'Area Of Interest' as well as the 'Anchor Feature Box'.  We now need to resize and position the AOI by dragging the control handles at its corners.  If the AOI is made too large then it is likely that lots of frames will be discarded as few frames will contain the complete AOI.

This is the Image Preview window after the AOI has been resized and positioned to select the part of the frame we wish to appear in the final video.

Unchecking the 'Show AFB/AOI' control shows just the output frame as it will be in the output video.

Looking back at the 'Processing Options' tab we see that the action of setting the AOI in the Image Preview window has updated the details in the 'Area Of Interest' control group.  The same was true for the 'Anchor Feature Block' when it was set in the Image Preview window.

Now we want to generate the a stabilised video using the selected options by going to the 'Do Processing' tab and clicking the 'Start Processing' button.

As always we should check the Summary details in the output area.  In this case we see that from 1433 input frames, 44 frames have been discarded as before because the AFB could not be found  In addition to this, 664 extra frames have been discarded because they did not contain the complete AOI.  The upshot of this is that the output video only contains 725 frames, though this obviously depends on the input video and the AOI chosen.

This is the generated video:

Note that although this video is stabilised without the black bars at the cost of the extra frames being discarded.  This video is suitable for moving on to the stacking phase of processing.