I’m working on a project where I’m scanning rock features (cliffs) that tend to be occluded by trees. These cliffs can span for a few miles so I take images far away to get the large photogrammetry model and then I take images up close to get the details in the rock, unobstructed by surrounding trees. It works pretty well and I only need to use control points occasionally to create a cohesive mesh. The issue I’ve ran into is that when I go to create the texture, all the images are used to texturize a given feature, so there are branches/twigs which become part of the texture, despite having images that much better represent the texture of the rock being included. I don’t want to exclude the far away images from texture creation, because they still contribute to the much larger texture. I would like to 1) give a higher weighting to the images which encapsulate the features and/or 2) mask the trees out of the texture. Does anyone have any suggestions on how to go about this using RC or 3rd-party tools to include in the pipeline?
I’ve included 3 images: one of the model that shows the model texture with the branches on it, one close up with minimal trees obstruction, and one further away with trees obstructing the rock. I’ve included a green circle as a reference location and boxed a region where this is occurring. Note, there are multiple images up-close and far away, I’m just showing one example.
