How do I convert hexagonal-prism panorama faces into an equirectangular panorama?
Asked 1/31/2014
3 views
2 answers
0
I received 8 portrait images from an older virtual tour: 6 images form the sides of a hexagonal prism, plus 1 top and 1 bottom image. Each file is 2848×4288 px. The six side images appear to meet edge-to-edge with no overlap, and the top/bottom faces overlap those edges.
The original tour was built in Pano2VR, but it won’t accept these source files directly. Because the side images have no overlap, normal stitching doesn’t work.
Is this a recognized panorama format, and what’s the best way to convert these prism-face images into a standard equirectangular panorama?
Originally by Photography Stack Exchange contributor. Source · Licensed CC BY-SA 4.0
Photography Stack Exchange contributor
12y ago
2 Answers
6
You can try to add images into Hugin and calculate) coordinates of images (or move them in OpenGl preview by hand), six faces will differ Yaw (0,60,120,180,240,300), top and bottom will have Pitch +90 and -90 , and then you must figure right projection (switch objectives) and hfov to "fill the frame". Then you can render it.
I have successfully used this method to convert six cube faces into equirectangular panorama.
Originally by user4332. Source · Licensed CC BY-SA 4.0
user4332
12y ago
0
Generated from our catalog & community — verify before relying on it.
Yes—this is essentially a face-based panorama, similar in concept to cube faces, but arranged on a hexagonal prism.
Because the side images have no overlap, you usually won’t stitch them conventionally. Instead, place each image at the correct orientation in panorama software and render out an equirectangular result.
A practical approach is to use Hugin: load the 8 images, manually set their positions, and render. For the 6 side faces, set yaw values to 0, 60, 120, 180, 240, and 300 degrees. Set the top and bottom images to pitch +90 and -90 degrees. You’ll then need to choose the appropriate lens/projection settings and horizontal field of view so each image fills its intended face correctly, then export as equirectangular.
If that’s difficult, another workaround is to use a 3D tool such as Blender: map the images onto a hexagonal prism mesh, align a camera, and render a panorama or intermediary views for stitching.
So the format is usable, but it requires manual geometric placement rather than automatic stitching.
Recommended products
UniqueBot
AI12y ago
Your Answer
Related Questions
Can Hugin convert six cube-face images into an equirectangular panorama?
Why do I get black holes at the top and bottom when converting a 360 cylindrical panorama to equirectangular or cubic in PTGui?
Could faint vertical bands on the edge of my Leica M6 negatives be a shutter issue rather than a light leak?
Why did some photos from my Fuji disposable camera come back green, hazy, and unusable?