PHOTOGRAMMETRIC ENGINEERING & REMOTE SENSING
November 2015
831
Q: In the process of adopting the new ASPRS Positional
Accuracy Standards for Digital Geospatial Data for our
orthoimagery production process, I faced the challenge
of evaluating the relationship between the imagery pixel
ground resolution and the horizontal accuracy class
assigned to orthoimagery according to the new standard.
Could you please advise me on the best way to relate the
pixel size and product’s horizontal accuracy according to
the new ASPRS Map Accuracy Standard?
Dr. Srinivasan Dharmapuri, Michael Baker
International, Pittsburgh, PA, USA
Dr. Abdullah:
The new standard was intentionally
designed to be versatile, leaving technology to be independent
to accommodate all possible products from current and
future geospatial mapping advancements. For this reason,
the new standard is not based on product pixel resolution as
the acquisition parameters for imagery with a certain pixel
resolution may vary widely depending on sensor geometry
and other acquisition parameters. The new standard merely
defines the horizontal accuracy class of a product by the root-
mean-square error (RMSE) required for that product. For
example, if the user needs orthoimagery or a planimetric
map that is accurate to 15 cm as RMSE, the new standard
assigns an accuracy class that is called “15-cm.” The class is
independent of the pixel resolution of the raw imagery or the
orthoimagery. This 15-cm accuracy class could be assigned to
orthoimagery with ground sampling distances (GSD) of 5 cm,
7.5 cm or 15 cm if the end user and the data provider mutually
agreed on such specifications. Of course, using imagery with
a GSD of 15 cm will require a more stringent workflow, very
accurate ground control points and airborne GPS. During
the past two decades and since we transitioned to softcopy
photogrammetry using either scanned film or digital camera
products, we adopted a common practice to relate the imagery
ground pixel resolution to the accuracy figure that is assigned
to a map scale according to the legacy ASPRS map accuracy
standard of 1990. During this practice, we rightfully—or
wrongfully—translated (associated) the map scale of the legacy
standard to an accuracy figure in terms of pixel resolution. It
became an industry standard to use the association of scale,
pixel resolution and map accuracy presented. This association
is demonstrated in Table 1.
Table 1: Relationship between pixel size and derived map scale.
Imagery GSD 3 in / 7.5 cm 6 in / 15 cm 1.0 ft / 30 cm
Supported C.I.
1.0 ft / 30 cm 2.0 ft / 60 cm 4 ft / 1.2 m
Supported Map Scale
1 in = 50 ft
1 in = 100 ft
1 in = 200 ft
1:600
1:1,200
1:2,400
Class 1 Ortho
Accuracy (RMSExy)
15 cm or 2
pixels
30 cm or 2
pixels
60 cm or 2
pixels
Table 1 reflects the following relationship between the pixel
resolution and the map accuracy that we used in specifying
map products horizontal accuracy until recently:
Class 1 Horizontal Accuracy of Map or Ortho is
RMSE
x
or RMSE
y
= 2 x Pixel Size (or GSD)
The new ASPRS standard includes an informative section
(section A.2) that describes the historical evolution of the
common practice of associating pixel size and accuracy. It
discusses why we should abandon the use of such practice.
Although informative, I believe that the previous statements
did not address your concerns, and they did not provide you
with the needed advice to relate product accuracy to the
imagery ground resolution according to the new ASPRS
accuracy standard. To do so, I would like to bring your attention
“The new standard was intentionally
designed to be versatile, leaving
technology to be independent to
accommodate all possible products from
current and future geospatial mapping
advancements”
Photogrammetric Engineering & Remote Sensing
Vol. 81, No. 11, November 2015, pp. 831–832.
0099-1112/15/831–832
© 2015 American Society for Photogrammetry
and Remote Sensing
doi: 10.14358/PERS.81.11.831
