Vision-guided servoing with feature-based trajectory generation (for robots)

Abstract

Absstract-This paper presents a vision module which is able toguide an eye-in-hand robot through general servoing and tracking problems using off-the-shelf image processing equipment. The vision module uses the location of binary image features from a camera on the robot's endeffector to control the position and one degree of orientation of the robot manipulator. A unique feature-based trajectory generator provides smooth motion between the actual image features and the desired image features even with asynchronous and discontinuous vision updates. By performing the trajectory generation in image feature space, image processing constraints such as the feature extraction time may be accounted for when determining the appropriate segmentation and acceleration times of the trajectory. Experimental results of a PUMA robot tracking objects with vision feedback are discussed.

Skill Transfer Between Industrial Robots by Learning from Demonstration

“…We used it to guide the HP3JC. There are two main approaches to visual servoing: Position Based Visual Servo (PBVS) [40][41][42] and Image Based Visual Servo (IBVS) [43][44][45][46][47][48]. Generally speaking, PBVS uses visual data to reconstruct the 3D world and allows researchers to design control algorithms in Cartesian space.…”

Position-Sensitive Detector (PSD) Guided Servoing Method for Industrial Robot Calibration

“…Most experimental systems reported in the literature operate at relatively low sample rates, and have low resolution or small field of view (FOV). An explicit trajectory synthesis algorithm is often employed due to the low visual sample rates (Feddema and Mitchell 1989). Many of these systems have to use low control gains for stable and low speed closed-loop motion.…”

Designing and Building Controllers for 3D Visual Servoing Applications under a Modular Scheme

“…Notice that visual servo structures require the direct access to the robot operating system in order to govern the joint variables directly, whereas dynamic look and move structures consider the robot as an independent system and can handle it as a black box. There has been a significant amount of research activity on image based control methods (Weiss et al, 1987), (Feddema & Mitchell, 1989), (Chaumette et al, 1991), (Espiau et al, 1992), (Khosla et al, 1993), (Corke, 1993) whereas there have been only a few researchers working on position based control methods (Koivo & Houshangi, 1991), (Allen et al, 1993), (Wilson et al, 1996), (Vargas et al, 1999). This tendency can be justified because image based systems usually reduce computation delays and eliminate errors due to sensor modeling and camera calibration processes.…”