In ROS I can publish a tf of a camera frame and a point cloud and have it build a map. But what if I just wanted to publish the point cloud. Can I adjust the xzy for each point in the point cloud message relative to the camera frame pose. Does anyone know how to do this?

I am thinking maybe this is what I want: where transform is the pose.

void pclros::transformPointCloud ( const pcl::PointCloud< PointT > & cloudin, pcl::PointCloud< PointT > & cloud_out, const tf::Transform & transform )