X hits on this document

PDF document

A Prototype Optical Tracking System Investigation and Development - page 30 / 170

376 views

0 shares

0 downloads

0 comments

30 / 170

16

Design Overview

but to produce a system capable of a high resolution.

The Fastrak resolution is defined in the manual [9] . The positional resolution decreases with range. The resolution at a distance of 1 cm is approximately 5 µm RMS. The angular resolution is quoted at 0.025RMS. The goal is to achieve resolutions of at least 5 µm and 0.025RMS over the entire tracking volume and allow future calibration to achieve the accuracy of the Fastrak or better. The Fastrak has a positional accuracy of 0.8 mm and 0.15angular accuracy. In this research resolution is quoted in terms of standard deviations instead of root-mean-square measures (RMS) that the Fastrak manual uses. Both measures are interchangeable.

Update rate

The Polhemus tracker achieves an update rate of between 30 and 120 Hz depending on the number of receiver units in use. In a standard one-receiver configuration the rate is 120 Hz. This is the rate at which pose estimates are generated. The update rate affects the speed at which an object can be scanned. greater update rate means a faster scanning speed can be used to produce the same quality scanned data. Initial component identification revealed that many image sensors do not support rates as high as 120 frames / s. It was determined that an update rate of 60 Hz was likely to be fast enough. This is justified in Chapter 3. If this number needs to be increased in the future there are options available to do this. These are firstly to use an image sensor that supports higher frame rates and

secondly to synthetically increase the update rate.

The

image

sensor

frame

rate

does

not

necessarily

equate

to

the

system

update

rate.

There

are two options to increase the update rate without changing the frame rate.

s

be a number of camera modules in use at a given time,

the first option is to

clocks belonging to the modules’ image sensors with respect to each other.

By

there will offset the offsetting

the update point for each module the update rate could effectively be increased by a factor equal to the number of camera modules present in the system. For example, for three modules the update rate could be lifted from 60 Hz to 180 Hz. The second method is similar to the first but involves calculating a pose estimate as soon as a marker location is updated by a camera module. For example, if one camera module is tracking 10 markers then this could potentially increase the update rate by 10 times. Care would need to be taken to ensure that the pose estimation algorithm worked correctly with marker positions

that

could

come

from

different

instants.

Tracking latency

Latency measures the time it takes for data at the input to the system to become available at the output of the system. This is the responsiveness of the system and affects the user

Document info
Document views376
Page views376
Page last viewedTue Dec 06 14:28:16 UTC 2016
Pages170
Paragraphs6307
Words54996

Comments