Analysis of Sensing Modalities for Electrode-Induction Gas Atomization of Metal Powders: Event Camera Data (RAW)

<h2>Event Camera Data</h2><p dir="ltr">Recorded event camera footage for the atomiser experiments performed as part of the "Analysis of Sensing Modalities for Electrode-Induction Gas Atomization of Metal Powders" paper.</p><p dir="ltr">The camera was positioned looking into the main atomising chamber. At the top of the chamber the molten metal drips down and hits a gas jet which forces the powder apart. This creates a large flash followed by objects seen going passed the camera. The event camera was chosen to try and capture the high moving objects going by as they travel towards the collection chamber below.</p><p dir="ltr">The files are compressed event camera data files organised into folders for each of the data runs. The data files are <a href="https://docs.prophesee.ai/stable/data/encoding_formats/evt3.html" rel="noreferrer" target="_blank">EVT3</a> data files.</p><p dir="ltr"><b>Author's Note: Run 1 had to be restarted due to the number of encoding errors causing the program to crash.</b></p><p dir="ltr">The RAW files can be opened and inspected either using Metavision Studio or the code samples for building your own viewer, both of which require installing the <a href="https://support.prophesee.ai/portal/en/kb/prophesee-1/metavision-sdk" rel="noreferrer" target="_blank">Metavision SDK</a>. It can be downloaded <a href="https://www.prophesee-cn.com/metavision-intelligence-sdk-download/" rel="noreferrer" target="_blank">here</a>. It is recommended to use Metavision Studio as a starting point since it has many useful tools such as scrubbing the timeline, different colour maps, exporting to video and trialling different accumulation times.</p><p dir="ltr">Metavision has a <a href="https://docs.prophesee.ai/stable/samples/modules/stream/file_to_hdf5.html" rel="noreferrer" target="_blank">tool</a> for converting it to HDF5 files but in this process, the invalid frames are dropped which may be of use hence uploading the full raw datasets.</p><p dir="ltr"><b>Authors note: The resulting HDF5 files did have invalid header data on some occasions so may be difficult to read in certain programs, depending on how they deal with it.</b></p>