Photogrammetry Test 4 - Outdoor Nike Jordan
Photogrammetry Outdoor Shooting
So today the weather finally cleared up enough to where I could attempt some outside shooting to see if I could yield better results with the photogrammetry and try to recreate some of the processes I have seen in the many youtube tutorials I have watched on the subject.
I opted for the Nike Jordan shoe again just as I already have an reference of how it can go completely wrong and so I wanted to be able to compare the results through this next experimental phase.
Unfortunately since I was on a time frame while I still had the light on my side I didn't get a shot of the setup, however below are some of the raw photos I took to give an idea of how it was setup.
I opted for the Nike Jordan shoe again just as I already have an reference of how it can go completely wrong and so I wanted to be able to compare the results through this next experimental phase.
Unfortunately since I was on a time frame while I still had the light on my side I didn't get a shot of the setup, however below are some of the raw photos I took to give an idea of how it was setup.
I set the shoe up on my camera tripod, which in hindsight was not the best shooting platform but it sufficed enough as I just needed to elevate the shoe off the ground. As you can see from the above images the lighting was pretty optimal for an outdoor shoot as it was completely overcast.
This first shoot I had around 70 photos and tried to take a variety of angles, but the one thing I struggled with was the bottom of the shoe and capturing any detail there.
Once I had finished shooting I went to my computer to export the photos into Reality Capture in order to see what result I would get.
This first shoot I had around 70 photos and tried to take a variety of angles, but the one thing I struggled with was the bottom of the shoe and capturing any detail there.
Once I had finished shooting I went to my computer to export the photos into Reality Capture in order to see what result I would get.
While there are some glaring issues that are quite obvious from first glance, I am still happy with this test. Previously when shooting indoors I only got half of the shoe which isn't ideal, however this time I managed to get a full model of shoe, but the issue being was the quality of the top of the shoe, I assume this was based on a lack of images captured which is better to note for next time.
I wasn't too worried about the underside of the shoe as typically you wouldn't see these in a game, unless the shoe is upturned for some reason, as such this can be ignored somewhat unless a full complete render is needed.
After trying these photos out I decided to auto expose the images I captured in Adobe Lightroom, to see if it was just a case of lighting as well, I then reimported these into Reality Capture and attempted another model. However the results were only marginally different and not enough to warrant exporting the model yet.

Thankfully I have kept all of the photos on my computer so I may try to reuse these at a later point or with more knowledge I may be able to fix the issues with the render and be able to finish this as a model.
I did attempt a second reshoot while there was still light and this time I tried to overshoot and take as many pictures as I could, I also overturned the shoe to shoot the underside properly. However I ran into issues when importing them into reality capture as the images were far too dark due to the lighting conditions having changed since I did the first shoot. Again if I improve my knowledge and understanding I may be able to utilise the best images from both shoots in order to create a better result.
Below are some images from the second shoot;
In total I have around 200 photos from the outdoor shoot so I'm sure I can use these in the future. Even if just as modelling reference.






Comments
Post a Comment