Hello,
This work is truly significant. I am currently researching this area and am curious about one aspect. If we are using known point cloud data and have obtained scene images from different sources, possibly taken with smartphones, it seems important in this method to determine the pose of the camera that took the images within the point cloud coordinate system. This likely involves image-to-point cloud registration. However, it appears that this aspect is not mentioned in the project. How did you handle this in your implementation?
Thank you!
Hello,
This work is truly significant. I am currently researching this area and am curious about one aspect. If we are using known point cloud data and have obtained scene images from different sources, possibly taken with smartphones, it seems important in this method to determine the pose of the camera that took the images within the point cloud coordinate system. This likely involves image-to-point cloud registration. However, it appears that this aspect is not mentioned in the project. How did you handle this in your implementation?
Thank you!