Getting images to align correctly by adding control points in reality scan

Hi, I’ve been usingReality Scan for a while and I’m 3d scanning a Old Bulky Organ which has giving me troubles, I first masked all images, and then got most images to align but specifically on the back there is no cameras aligned . The problem it’s mostly a flat surface without much detail other than some screws and breathing holes.

I’m not familiar with adding control points, but I would like to be able to manually add some. The add control control points option in the alignment tab only lets me click around where the existing point cloud is made which doesn’t have details from the missing photos In the image you can see the back is completely missing. In total I have 365 photos and only around 160 got aligned. I could just try to adjust alignment settings but I think forcing it would cause even worse results.

I did realize I took pictures of the back very close so you can’t really see the background, which is probably the reason it has failed.. I would like to share the dataset but it contains pictures of my family In the background… Thanks

Hi, there are multiple ways to add control points. One thing you already tried was adding them to the 3D viewport. For that, you already need some alignment.

There is another way: place control points directly in images, even in unaligned ones. Simply open an image in a 2D VIEW, activate the control point tool, and place it on a recognizable feature, which you can identify in neighboring images. Mark the same control point in multiple images. Repeat this for multiple control points so RS has enough information to make the 3D reconstruction.

This will be difficult and might not work in the end. It is only worth trying out if you don’t have an option to take more photos.