X hits on this document

PDF document

AIRBORNE KINEMATIC GPS POSITIONING FOR PHOTOGRAMMETRY THE DETERMINATION OF THE CAMERA EXPOSURE ... - page 3 / 11

24 views

0 shares

0 downloads

0 comments

3 / 11

time. (time

As mentioned above, the length of the between successive epochs of data) is

uniform interval governed by the

GPS receiver hardware and accurate interpolation of the aircraft trajectory.

may be the limiting factor for the the exposure station position from The interpolation precision is well

correlated to the limit of positional expected from kinematic GPS (Lapine, intervals degrade the interpolation.

resolution which can be

1991). Longer

time

The following

example

used GPS data collected at a 1-second rate to 2-second and 5-second rates by removing

sample points.

The thinned data sets

the

missing

midpoint

samples.

The

GPS

were data

and then thinned the appropriate interpolated for used for this test

came from two data sets collected aboard the jet and one data set from the Texas Highway

NOAA Citation II Department King

Air Turbo-prop.

Comparison of the interpolated positions with

the

observed

positions

indicate

a

standard

deviation

about

the

mean

difference

as

great

as

41

cm.

The

number

of

samples

whose

position difference was greater than 20 centimeters was recorded

for

each

data

set

and

reported

in

Table

1.

with

The Citation and King Air results are still commensurate kinematic positioning expectations when thinned to 2

seconds.

The sample standard deviations for the

to 5 seconds in position. the standard

for the King Air suggest a serious It is interesting to note that the deviations associated with the NOAA

data thinned degradation magnitude of Citation are

different from

the Texas

were tested and

failed the

King Air. The population equality test based on the F

variances statistic

(Hamilton

1964).

The

differences

between

the

2-

and

5-second

populations are most likely caused by the aircraft trajectories over a time

the inability to model span greater than 2

seconds.

The

difference

between

the

NOAA

and

Texas

populations

may

result

from

the

same

inability

to

model

the

trajectories

or may be receivers

caused by a larger used for the tests

signal-to-noise ratio in the

(different

manufactures).

The

trend

indicates

that

sampling

intervals

greater

than

1

second

should be avoided if the positioning is required.

full accuracy of kinematic GPS However, the good comparison between

the data thinned to validate the ability at 1 hertz.

2 seconds and observed positions does of the interpolation process when sampling

An

interesting

alternative

to

the

above

timing

situation

would be to activate the camera shutter pulse generated by the GPS receiver. may be able to entirely eliminate the

with the 1-second timing In this procedure one time difference between

GPS fix information and exposure.

navigation

file

would

be

eliminated

Interpolation of

except

possibly

the for

GPS a

constant response aircraft

time offset between the timing signal and the camera

to the signal. indicated that

A short test conducted aboard the NOAA the time delay would be on the order

of 0.1 test.

second for the particular Wild RC-10 camera used in the The camera service manual states that the maximum time

delay between rotating shutter blade opening and capping shutter

delay

is

on

the

order

of

0.070

second.

The

time

delay

depends

on at

the the

shutter speed, position of the time the pulse was initiated,

rotating shutter blades and the vacuum status.

One major problem with this procedure to accurately control overlap.

would

be

the

inability

Document info
Document views24
Page views24
Page last viewedWed Dec 07 14:49:54 UTC 2016
Pages11
Paragraphs1280
Words4188

Comments