Although excellent at making identical repetitive movements, robots lack perceptual ability for objects placed at random locations. Now, researchers aim to provide new solutions to robotic vision enhancement by creating a filter in a 6D pose versus a 3D. The studies were based on improving computer vision and developing such filter will give robots an accurate spatial perception and aid them in navigation and manipulation of objects.
University of Illinois at Urbana-Champaign: Overview of the PoseRBPF framework for 6D object pose tracking. The method leverages a Rao-Blackwellized particle filter and an auto-encoder network to estimate the 3D translation and a full distribution of the 3D rotation of a target object from a video sequence. Image Credit: aerospace.illinois.edu
The 6D pose gives robot a more complete picture of the relative location of an object in respect to the camera—it’s worth noting that although a 3D pose does give location information on X, Y, and Z axes, it is not as accurate.
"Much like describing an airplane in flight, the robot also needs to know the three dimensions of the object's orientation -- its yaw, pitch, and roll," said Xinke Deng, doctoral student studying with Timothy Bretl, an associate professor in the Dept. of Aerospace Engineering at U of I. "We want a robot to keep tracking an object as it moves from one location to another.”
In real-life stimulation, the dimensions of the 6D are in constant change and thus the filter, which looks at particle location, helps robots analyze spatial data.
"In an image-based 6D pose estimation framework, a particle filter uses a lot of samples to estimate the position and orientation," Deng said. "Every particle is like a hypothesis, a guess about the position and orientation that we want to estimate. The particle filter uses observation to compute the value of importance of the information from the other particles. The filter eliminates the incorrect estimations.”
Watch this video demonstration to learn more:
"Our program can estimate not just a single pose but also the uncertainty distribution of the orientation of an object," Deng said. "Previously, there hasn't been a system to estimate the full distribution of the orientation of the object. This gives important uncertainty information for robot manipulation. Our approach achieves state-of-the-art results on two 6D pose estimation benchmarks.”
Source: Science Daily
