Junkers Ju 52 Photogrammetry Model (Kea, Greece — 205 fsw)

Background

This is the second post for the Junkers Ju52 Photogrammetry project. The first post covered the trip details, information on the Junkers Ju52 airplane, the wreck itself, and still photos of the wreck. This post will focus on the photogrammetry model.

Photogrammetry Process Photos & Video

One of the cool things about this project is that Ben & Lauren were with me so Ben was able to take a bunch of photos and made a video of me in the process of taking photos for the model. I took about 2,200 photos across two dives (details in the last section of the post) while Ben took stills and video.

Photogrammetry Process Photos

Below are a few of the photos Ben extracted from the video he shot. I need to cover every inch of the wreck top to bottom to get the images for the photogrammetry model.

Photogrammetry Process Video

Ben also shot some cool video of the wreck which also shows some of the process of taking the photos for the model (try to ignore my trim!).

Online 3D Model Links

Below are links to two different versions of the online 3D photogrammetry model. One is better suited for mobile devices and requires less memory. The other one might work on some mobile devices but will definitely work on computers.

Mobile Devices (lower resolution)

Desktop Version (higher resolution)

Model Screen Captures

Below are screenshots of the full model from multiple different angles. I highly recommend checking out the on-line 3D model, but these will give you and idea of the wreck even if you don’t.

Model Details

Given the amount of photos and the time taken, there are some really nice details in this model which I wanted to highlight.

The first is the “see through” glass in the cabin. If you look at the online model, you can see the seabed through the broken windows as you move the model around. It came out very nice. Below is a static screen capture.

I also took pictures inside the cabin and inside the gunner spot near the tail so you can view inside the wreck in some amount of detail. If you look closely at the screen capture in the middle left, you can see the intact glass on the port side of the cabin with some growth on it.

Noticing that the vertical rudder was missing after building the model on the first dive, I made sure to go back and get photos to tie it into the overall model on the second dive. It is a shame that it was knocked off at some point. Also note how the white downline is gone from the starboard side stabilizer. 🙂

Finally, the three engines and the cockpit area. There is a little bit of detail lacking on the engines which I wish were a bit better, but they did turn out well.

Building the Mode (Details)

For those readers who want more details on how the model was built, I’ve provided a summary of progress across both dives along with some interesting data about the model below.

Dive 1

Dive 1 was on 16 April and had a max depth of 206 feet with about 35 minutes of bottom time and about 80 minutes of decompression. The plane is a pretty good size (96 foot wingspan, 62 foot length) but I was able to get coverage of most areas and took 1365 images for the model on the first dive. I had my laptop so I built a low-res model while in Kea to ensure that the photos aligned and to see areas I missed. Below are some screen captures of them “Medium” resolution model after Dive 1.

I have provided three views. The first is the model, the second shows placement of the photos and the third shows “confidence” of the model. Blue is high confidence, green is medium and red is low.

Based on just photos, it looks like pretty good coverage. However, you will note in the confidence view that there are some red areas under the wings and horizontal stabilizers and also near the engines/props at the front. Under the wings are a bit tough and not really as important; however, the area near the props definitely needed more coverage and photos on the second dive.

We had one other “problem” with the first dive model. The downline was draped on top of the starboard side stabilizer which was not a good look and not something I wanted in the final model!

Dive 2

I was lucky to have Ben & Lauren with me and they moved the downline off the wreck so that I would be able to re-shoot that part of the wreck for the final model.

The second dive was on 18 April and was a max depth of 210 feet with a bottom time of about 25 minutes and about 60 minutes of deco. I shot an additional 865 photos and concentrated on four areas: (1) the props, (2) the areas under the wings and the horizontal stabilizers, (3) the starboard side horizontal stabilizer and (4) inside the rear gunner position (I had covered most of the cabin area on the first dive).

After the second dive, I built a model on-site to make sure I got what I needed. Below is a “front on” view showing the increased details of the props and areas under the wing and stabilizers in the rear. In fact, you can see all the way through to the ocean which is what I wanted. The confidence model shows a little bit of red, but definitely mostly covered. The port side engine could have used a little more details but it will have to work!

In terms of the starboard stabilizer, after the second dive, the downline wasn’t completely gone since it is an amalgamation of both dives. However, when I build the final model, I will omit photos from that area from dive 1.

Coloring

The model as shot originally definitely has a “blue” tint. I’ve spent some time trying to do some color correction which is NOT easy given that there are over 2,000 photos! I’ve gone through two iterations of the model with different sets of photos. I’d be interested to get any feedback on which one is the most appealing.

The interesting aspect is that the third one “second color correction” is probably closest to the actual colors of the airplane. For example, here is sample photo straight out of the camera:

In the end, I went with the “as shot” model for historical purposes but I will probably “play around” with other options over time.

Thanks To

These projects are always resource intensive and take a lot of time and effort to coordinate and pull off. A number of people deserve recognition:

• Ben Lair & Lauren Lasser were great to dive with and I’m so happy they were already in Europe and able to make the trip down to Greece to join up with me. It worked perfectly.
• Yannis, Tasos, and the crew @ Keadivers. As usual, they are true professionals and it was great to spend some time in the water with Yannis. He was instrumental in obtaining the clearance to dive the wreck. Always a pleasure!
• The Greek government was very accommodating and I’m appreciative to have a chance to dive the Ju52 and to document it with a photogrammetry model.
• Russ Matthews at Air/Sea Heritage Foundation was critical to get support from an archaeological concern for the Greek government permit process.
• My wife had to listen to me incessantly talking about diving this “Ju52 airplane” in Greece and then, of course, not even knowing if I was going to actually travel based upon our permit status. Thanks, baby!

MISSION ACCOMOPLISHED!