Towed Camera Array for underwater Photogrammetry on a budget . . .

[Scott McWilliam]

Underwater Photogrammetry

 

[Johnoly]

My corporate computer won't let me D/L the PDF, but I did attend a great seminar at DEMA19 three weeks ago on Photogrammetry. I've seen on the internet 3D renderings of sunken ships and always thought they were very interesting but never knew the amount of work on how they were made. I forgot the presenter's name but he did a great job explaining how he photograph's a wreck sometimes over a couple of dives or with a couple of divers and up to 1000 photos. He then used a beefy desktop computer that took photo number 1 and it compared it to photo number 2. It identified and matched as many different pixels in each photo and then linked them together even including vectors and distances. That desktop computer continued to take each of the 1000 pictures and compare each one to all the others, pixel by pixel and vector by vector. He said on one wreck, after he set it up his computer and let it start crunching that it took 11 days just to get the data points, not even to fill in the colors and solid space between each point.

It was a very cool presentation and I definitely have new appreciation for the handful of people that are doing this to record wrecks and other underwater objects. Seminars like this are the 2nd best part of attending DEMA !!

 

[Scott McWilliam]

It may have been the same unit.

 

[johnkendall]

My corporate computer won't let me D/L the PDF, but I did attend a great seminar at DEMA19 three weeks ago on Photogrammetry. I've seen on the internet 3D renderings of sunken ships and always thought they were very interesting but never knew the amount of work on how they were made. I forgot the presenter's name but he did a great job explaining how he photograph's a wreck sometimes over a couple of dives or with a couple of divers and up to 1000 photos. He then used a beefy desktop computer that took photo number 1 and it compared it to photo number 2. It identified and matched as many different pixels in each photo and then linked them together even including vectors and distances. That desktop computer continued to take each of the 1000 pictures and compare each one to all the others, pixel by pixel and vector by vector. He said on one wreck, after he set it up his computer and let it start crunching that it took 11 days just to get the data points, not even to fill in the colors and solid space between each point.

It was a very cool presentation and I definitely have new appreciation for the handful of people that are doing this to record wrecks and other underwater objects. Seminars like this are the 2nd best part of attending DEMA !!

That was probably me I don't recall there being any other presentations about Photogrammetry, however I could well be wrong. You can see some of our work here: www.sketchfab.com/GUE

Thanks
John

 

[Scott McWilliam]

Lol, no your probably right I was not there. I think it was DEMA funding that launched the Park Service development of their system.

 

[Scott McWilliam]

Better link
Underwater Photogrammetry