Deep EndoVO: A Recurrent Convolutional Neural Network (RCNN) based Visual Odometry Approach for Endoscopic Capsule Robots.
Mehmet TuranYasin AlmaliogluHelder AraújoEnder KonukogluMetin SittiPublished in: CoRR (2017)
Keyphrases
- convolutional neural network
- visual odometry
- autonomous navigation
- mobile robot
- face detection
- long range
- simultaneous localization and mapping
- kalman filtering
- ego motion
- multi robot
- depth images
- robotic systems
- autonomous robots
- path planning
- position information
- endoscopic images
- neural network
- kalman filter
- object detection
- pairwise
- extended kalman filter
- real robot
- indoor environments
- range data
- dynamic environments
- super resolution
- detection method