Visual Servoing

  • Visual Servoing with Safe Interaction:  The control of the interaction during an image based/position based visual servoing for a robot working in dynamic cluttered environments is considered. The main concerns in this scenario are the performance of the main visual servoing task, keeping the visual feature in the field of view as well as a safe physical    avoidance/interaction.


  • Pose Estimation using Deep Learning Approaches:  A new Convolutional Neural Network is studied for the purpose of robot pose estimation. The model does not require any knowledge about the camera intrinsic parameters. New dataset is introduced and precise results are achieved via the proposed CNN model during the experiments.

Dr. Hamid Sadeghian, Assistant Professor, Engineering Department, University of Isfahan, Isfahan
Dr. Zahra Kamranian, Postdoctorial Fellow, Computer Engineering Department, University of Isfahan, Isfahan
Dr. Abbas Karami, Phd, Mechanical Engineering Department, Isfahan University of Technology, Isfahan.
Related Publications:

  1. Hamid Sadeghian, and Luigi Villani, “Visual Servoing with Safe Interaction”, 8th International Workshop on Human-Friendly Robotics (HFR 2015), Munich, Germany, 2015.
  2. Hamid Sadeghian, Luigi Villani, Zahra Kamranian, Abbas Karami, “Visual Servoing with Safe Interaction using Image Moments”, International Conference on Intelligent Robots and Systems, IROS2015, Hamburg, Germany, 2015.