Question about the input of libviso2

edit

hi everybody.I am working on the project about using a single camera for navigation which will use the libviso2 to compute monocular odometry. I had read the post :http://answers.ros.org/question/46018/do-i-have-to-rectifycrop-images-before-process-them-with-viso2_ros-mono_odometry/ ,http://wiki.ros.org/image_proc and know that the input image should be rectified before feed to libviso2. What does the 'rectify' mean? Does it mean undistort image or anything else? I want to know more details about this word 'rectify' mean.Thank you everybody.