Request-and-respond approach to reducing latency within a...

Details Request-and-respond approach to reducing latency within a... Request-and-respond approach to reducing latency within a...

1999-06-22

2001-05-22

Nguyen, Chau (Department: 2663)

Surgery

Diagnostic testing

Measuring anatomical characteristic or force applied to or...

Reexamination Certificate

active

06234983

ABSTRACT:

TECHNICAL FIELD
The invention relates generally to motion tracking systems and more particularly to managing latency in a system for tracking movement of an object, such as a human head.
DESCRIPTION OF THE RELATED ART
There are a wide variety of applications in which it is desirable to rapidly and continuously determine the location and orientation of at least one object. Applications range from enabling virtual reality in a pilot training or game environment to enabling low-invasive surgery. In each of these uses, a high degree of sophistication is needed in designing system components, such as sensor circuitry for generating raw data and sensor processing circuitry for manipulating the raw data to extract location and orientation information.
Techniques for providing the raw data include utilizing multiple light sources and detectors that enable optical tracking by triangulation, using electromagnetic devices that generate electrical signals, and using magnetic sensors that are responsive to the magnetic field of the earth. U.S. Pat. No. 5,373,857 to Travers et al. describes a head tracking apparatus for a virtual reality system that tracks the movement of a person's head based upon a magnetic sensor that is embedded in a helmet and that is responsive to the earth's magnetic field.
In a system for tracking movement of an object, total latency is determined by the sum of the latency imposed by a tracking subsystem and the latency imposed by a display subsystem. The latency imposed by the tracking subsystem is determined by the physical characteristics of the sensors that generate the raw data, the delays introduced by the signal processing required to extract the appropriate angular and positional data from the raw data, and the transport delays required to transfer the angular and positional data to the display subsystem. This tracking latency is defined as the delay between a stimulus and the time at which the data pertaining to the stimulus is received by the processor of the display subsystem. In a head tracking system, the stimulus is in a form of head movement and the display subsystem generates image data for a trackable display device, such as a helmet-mounted display. The latencies of the display subsystem result from the finite time taken by the processor to render each new scene, using fresh head orientation and positional data from the tracking subsystem.
For a virtual reality application, if the sum of all latencies from the original head motion to the final displayed image is too great, there is a strong tendency that the user will suffer from some degree of “simulator sickness.” Simulator sickness is caused by the conflict in the person's cognition between its own head positional sensing and the virtual reality view that is presented at the display device. While simulator sickness is a particular concern to head tracking systems, minimizing latency is also a goal in the design of other types of tracking systems, such as those that monitor the position and orientation of a human hand.
U.S. Pat. No. 5,592,401 to Kramer describes a method of managing delays in a virtual reality system. By employing a combination of position-sensing technologies, position information from the combination provides a less delayed representation without an unacceptable reduction in the accurate reporting of movement of an object. Specifically, more than one sensor is used, with each sensor having different deficiencies and proficiencies. A first sensor may be highly accurate, but may provide a delayed signal with a latency that is unacceptable in many applications. A second sensor technology may provide high speed results, but with a susceptibility to drift and other inaccuracies. The combination of signals may be used to obtain high resolution, real time depictions of movement.
The delay of packets having positional data in a virtual reality system is also considered in U.S. Pat. No. 5,793,382 to Yerazunis et al. The patent is particularly concerned with providing smooth motion in a shared distributed virtual reality system, since jitter is likely to occur as a result of lost or delayed packets in a network environment. Smoothing is achieved by providing redundant data in packets, so that if one packet is lost, adjacent packets contain sufficient information to smooth the perception of displayed motion. As another technique, extrapolation is used in position, axis of rotation and angle of rotation, in the event of large amounts of missing data. Moreover, the information in the packets is categorized as to the predictability of future motion, and appropriate smoothing algorithms are applied based upon the predicted knowledge of what the future motion is likely to be. The system switches between a linear interpolation algorithm for less predictable data and a Catmull-Rom spline for more predictable data.
In addition to the previously recognized causes of disorientation and reduced perceived quality in a tracking system, it has been determined that variability in latency can lead to the perception of jitter. In this case, variability in the latency of the tracking data causes variable “step-sizes”between frames of the display. As a result, the perception of smooth movement is diminished. Variability in latency is primarily a result of asynchronous processing between the circuitry for generating the tracking data and the circuitry for driving the graphical display. Typically, tracking data is supplied at a fixed sample rate, while the time needed to render successive scenes is often scene-dependent.
Variable latency will be described with reference to FIG.
1
. Three times at which fresh tracking data is required for processing by a graphics driver are represented by points A, B and C. Since graphical processing requirements are scene-dependent, the spacing between time A and time B is different than the spacing between time B and time C. On the other hand, there are six sample times
10
,
12
,
14
,
16
,
18
and
20
which are equidistantly spaced to represent the capture of raw data from a sensor or sensors. The horizontal lines extending from the sampling periods
10
-
20
represent the scene-independent latency required for extracting positional and angular information from the raw data and for transporting the extracted information to the processor for a graphics driver. Typically, the graphics driver is contained in a host computer connected to the sensor processing by a data link which may be a wired or wireless connection.
As can be seen at time A, the fresh tracking data is coincident with the requirement for fresh tracking data, so the system latency is minimized. However, at time B and time C, the availability of the host for fresh tracking data is not coincident with the arrival of the data. At time B, the additional system latency is relatively small, since the supplied data from the sampling at point
14
arrived only shortly before time B. The latency associated with time C is significantly longer, since the last-supplied packet of tracking data was extracted from the sampling at point
16
. Because the host computer has no way of knowing that a short wait would enable it to access tracking data extracted from the sampling point
18
, the last-supplied packet of tracking data is utilized in the graphics processing. Thus, there is a Attorney Docket No. RET-001 different system latency for each of the three accesses of packets of tracking data at times A, B and C.
Increasing the sample rate of acquiring raw data leads to lower variability and the system latency experienced by the host computer, since the difference between the minimum latency and the maximum latency will always be less than one sample period. However, simply increasing the sample rate is not ideal from a system point of view. Firstly, the data must be transferred to the host computer via some form of data link. For ease of interfacing with most graphics hosts, this may be a serial link conforming to the RS-232C standard. Alternatively, for IBM-PC compatible machines, the link may include a

Affiliated with

Also associated with

Rating

Request-and-respond approach to reducing latency within a... does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Request-and-respond approach to reducing latency within a..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Request-and-respond approach to reducing latency within a... will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-2553771