Converting Kinect IR images to points
My goal is to compress a movie of kinect data into the smallest format possible. I'm investigating if I can use movie compression on either the raw /camera/ir/image_raw images or maybe the /camera/depth_registered/image_raw frames.
I'm using this tutorial to extract the frames from a rosbag recording. And it works great. For the next step, I have two questions:
Can you export image frames in png (i.e. a lossless) format for compression? I found this answer but I'd prefer not use openCV/python.
Is there a ros node to convert the raw image format into points. "convert_pointcloud_to_image" will convert points into images but I'm looking to use both the camera data & the image to get points & haven't found a tutorial or node for this...
I'm running Ubuntu 11.10 w/Electric & PCL 1.6, as I haven't had good luck with Openni drivers on 12.04. Looking in: /opt/ros/electric/stacks/perception_pcl/pcl_ros/bin I don't see anything that does what I would want. I assume this functionality exists somewhere (as the Openni drivers converts an inferred image into a point cloud) but I can't seem to find it.
Help please?
