X hits on this document

PDF document

A Prototype Optical Tracking System Investigation and Development - page 114 / 170

398 views

0 shares

0 downloads

0 comments

114 / 170

100

Results and Discussion

Background

Correspondence

Module Pose

Image

Algorithm

within Hub

Noise

Errors

Enclosure

Pose

Pose Estimate

Estimation

2 O

Marker Non- uniformity

Sensor Dark Pixel Noise

Centroid

xN

  • O

    1

Camera

Marker

Pose

Intensity

Ideal

Measured

Marker

Marker

Image

Image

Marker

Number

Number

Optimisation

Precision

Precision

Noise

Marker

3D

Camera

Manufacture

Marker

Intrinsic

Variation

Coordinates

Parameters

Constant unknown

Varying unknown

Figure 8.1 The uncertainties affecting different parts of the system are shown. Calibration can reduce the effect of constant unknowns. Modeling can reduce the effect of some varying unknowns. The remainder of the unknowns will reduce the accuracy of the final pose estimate. Point 1 shows the output of a Black Spot module. Point 2 is positioned at the output of the system.

on the markers’ current source and the electrical noise on the analog-to-digital converter in the image sensor. This type of unknown cannot be removed by calibration or modeling, and its effect will be seen at the output of the system.

The major constant unknowns are:

  • intrinsic camera parameters, i.e. focal length, principal point, pixel size, and sensor

skew,

  • 3D marker coordinates,

  • Black Spot module pose within the hub enclosure, and the

  • non-uniformity between markers.

The major varying unknowns are:

  • variation in marker intensity due to power source noise,

  • variation in marker intensity/marker size due to camera pose,

  • variation in centroid position due to background image noise,

Document info
Document views398
Page views398
Page last viewedWed Dec 07 22:33:29 UTC 2016
Pages170
Paragraphs6307
Words54996

Comments