Signal processing techniques for haptic data compression in teleoperation systems

In this paper, we present various signal processing techniques for haptic data compression in networked teleoperation systems or haptic interfaces. A novel rate-distortion optimized quantizer is introduced to compress haptic data whose dynamic range is limited. Prediction methods, which are primarily based on fixed rate down-sampling, are also presented and combined with the proposed quantizer for performance improvement. Experiments are performed using a haptic interface in a virtual environment to evaluate the proposed signal processing techniques. An objective performance measure, which is called the operational rate-distortion measure, is introduced to evaluate not only the compression ratio, but also the quality of reconstructed haptic data.