This is the third video in the series about 3D Reconstructions and in this video we’ll discuss the first half of Meshroom’s pipeline, we’ll talk about CCTags for automatic scaling, and we’ll talk about what changes when we use a Turntable. Finally, we’ll have a look at using known camera positions to improve the reliability of our reconstruction.
Originally I had planned to make a video about how I built the main part of the machine that I use to automatically take pictures. But there’s just so much to say that I didn’t really know where to start. It was a bit overwhelming, to be honest. So I’ve decided to skip that part (for now) and move on to describing how Meshroom works.
I think what sets my video apart from other videos about Meshroom is that I try to explain in a bit more detail what the different nodes do and why we need them. That gives you some insight into how the algorithms used by Meshroom work, and that will help you when it comes to figuring out why a particular set of pictures didn’t reconstruct.
In the next video we’ll investigate how the results of a reconstruction differ when we use pictures where the object is sharp from front to back as compared to when some parts are blurry. And as part of that video we’ll also look at focus stacking. We may also do a little investigation into quality vs quantity (of images).
Leave a Reply