hi , i worked with turtlebot2 and microsoft kinect, i want to evaluate pose estimation for three cases: odometry odometry+imu odometry+imu+visual odometry

and then integrate the best estimation in rgbdslam to estimate the pose, My first question: Is rgbdslam has visual odometry topic? because i didn't find this topic when i launch rgbdslam.launch, if not how can i obtain this topic? My second question: after compare these 3 estimations and find the best how to integrate it in rgbdslam? My final question: the trajectory estimate file provided by GUI interface in rgbdslam, how it was estimated ?

ANy help please

Thanks in advance