A Robust Vision Module for Humanoid Robotic Ping-Pong Game

Abstract

Developing a vision module for a humanoid ping-pong game is challenging due to the spin and the non-linear rebound of the ping-pong ball. In this paper, we present a robust predictive vision module to overcome these problems. The hardware of the vision module is composed of two stereo camera pairs with each pair detecting the 3D positions of the ball on one half of the ping-pong table. The software of the vision module divides the trajectory of the ball into four parts and uses the perceived trajectory in the first part to predict the other parts. In particular, the software of the vision module uses an aerodynamic model to predict the trajectories of the ball in the air and uses a novel non-linear rebound model to predict the change of the ball's motion during rebound. The average prediction error of our vision module at the ball returning point is less than 50 mm -a value small enough for standard sized ping-pong rackets. Its average processing speed is 120fps. The precision and efficiency of our vision module enables two humanoid robots to play ping-pong continuously for more than 200 rounds.

Real Time Trajectory Prediction Using Deep Conditional Generative Models

“…This physics model considers air drag and bouncing physics but ignores spin. To estimate the initial position and velocity, we use the approach proposed in [3], that consists on fitting a polynomial to the first n observations and evaluating the polynomial and its derivative in t = 0. We tried different values of n and chose the one with better results.…”

“…In addition, we introduce strategies to make the model robust to missing observations and outliers. We evaluate the proposed approach on a robot table tennis setup in simulation and in the real system, showing a higher prediction accuracy than a LSTM recurrent neural network [9] and a physics based model [3], while achieving real time execution performance. An open-source implementation of the method presented in this paper is provided [14].…”

Dynamic Nonprehensile Manipulation of a Moving Object Using a Batting Primitive

“…Chen et al [11] developed a vision module for humanoid robotic table tennis. The vision module contains two stereo vision sensors with a 200 fps and an algorithm for predicting the rebound trajectory of a table tennis ball.…”

Designation and Control of Landing Points for Competitive Robotic Table Tennis

“…As far as humanoid table tennis robot is concerned, the balance maintenance was discussed in paper [13]. We have also developed a humanoid robot which is able to rally with human players or another robot player, and the best record of cooperative playing (robot vs robot) is more than 200 rounds [14,15,16,17,18,19,8]. Most of the proposed robots were designed for cooperative table tennis playing, and have achieved a great success.…”