Under certain conditions, such as high winds or observing faint stars, it is advantageous to use the shortest exposure times (~20 msec) of the VisAO camera and perform Lucky imaging. This technique is based on first selecting only the best images in a sequence of short exposures, and then shifting-and-adding (SAA) the results. This process can be computationaly intensive. As such, we have implemented a pipeline for the VisAO camera which is based on Google’s Youtube service. First a sequence of frames is uploaded to YouTube. If SAA can be beneficial, YouTube will offer to fix it, stating that the video is “shaky”. Here we present a comparative analysis of this pipeline.
Before:
After:
As you can see, significant improvement is achieved!
Future work:
The next phase of implementing this data reduction pipeline will include direct framegrabber-to-Google uploading, preventing any need for local storage. This will be helpful in limiting the need for MagAO project personnel to cause so many problems at LCO.