example fiducial marker patterns

This page contains example fiducial marker patterns for different cameras that have acquired historical aerial photos, including what their fiducial marks look like, and the approximate coordinates of the fiducial marks that you can use to populate the MeasuresCamera.xml file.

Note

The fiducial markers in the following examples are labeled as follows, with the assumption that the data strip / direction of flight marker is on the left side of the image:

• P1: the lower left corner of the image

• P2: the upper right corner of the image

• P3: the upper left corner of the image

• P4: the lower right corner of the image

• P5: the middle of the left edge image

• P6: the middle of the right edge of the image

• P7: the middle of the top edge of the image

• P8: the middle of the bottom edge of the image

a note on identification

Aside from a single 1960 article[1], I have not managed to find much solid information about different fiducial marker patterns used by different camera manufacturers in different cameras. Some patterns are relatively distinct, but others are both very similar and widely used, making a definitive identification based on the fiducial pattern alone challenging, especially for earlier cameras that did not expose this ancillary information on the image negative.

A good deal of the information on this page has been pieced together from looking at a very large number of images and calibration reports downloaded from EarthExplorer[2] and the USGS Camera Calibration Report Database. This guide is by no means exhaustive, though it is built on a fairly large dataset of example images.

Note

If you have a calibration report that corresponds to your specific images, you should use that instead.

The information provided here is for those cases where a calibration report does not exist, or has been lost to time.

Tip

For images acquired by the US Government or other agencies and hosted on EarthExplorer, you may be able to find a calibration report by using the lens or camera serial number, if this was recorded on the camera’s data strip (or, sometimes, written around the edges of the frame).

With the serial number(s) and/or other identifying information (e.g., make/model), you can use the USGS Camera Calibration Report Database to search for the calibration report for your camera (or a similar camera).

Additionally, I have transcribed the information from a (growing) number of these calibration reports; the combined CSV files and a few jupyter notebooks for printing summary information based on make/model, fiducial marker pattern, etc. can be found at iamdonovan/fiducial_markers.

“wing-type” fiducials

A number of different camera manufacturers have used four “wing-shaped” mid-side fiducial markers of varying designs:

• Aero Service Corporation (also Aero/View cameras, or just “Aero”)

• Airagon

• Mark Hurd Aerial Surveys, Inc.

• Park Aerial Surveys, Inc.

• Fairchild

• Wes Smith

Many of the cameras made/used by Mark Hurd Aerial Surveys, Inc., Park Aerial Surveys, Inc., or similar companies do not appear to have model names - rather, they are identified only by a number. The lenses used in these cameras tend to be made by manufacturers such as Bausch and Lomb (e.g., the Metrogon Lens), C. P. Goerz (e.g., the Aerotar lens), or Zeiss (e.g., the Pleogon lens).

Tip

Because:

• the measured separation between P5-P6 and P7-P8 is somewhat variable for these images, even for the same camera manufacturer (and sometimes even the same camera as identified by the serial number…); and

• the majority of these cameras were used before coordinates were regularly included in calibration reports,

it is probably a better idea to use estimate_measures_camera with a large sample of these images (and the known scanning resolution), rather than the approximate coordinates given below.

hollow-type wing

Cameras with this style of fiducial marker (four mid-side ‘wing’ markers that are at least semi-transparent) seem to have been made by Mark Hurd Aerial Surveys, Inc. They do not have any particular model name, and are recorded with single- or double-digit serial numbers in most calibration reports on file. They used a few different lens types, mostly Bausch and Lomb Metrogon with a 6” (152 mm) focal length.

Marker Separation (n = 3 reports)

markers	mean	median
P5 - P6	221.887 ± 0.097	221.91
P7 - P8	221.780 ± 0.070	221.78

flat wing (small)

This style is perhaps the most common of the wing-type (at least by volume in EarthExplorer), and seems to have been used by a wide range of manufacturers (with the caveat noted above[2]):

• Fairchild (K-3B “modified”, K-17, K-17 “modified”, K-17B “modified”)[3]

• Aero Services (Aero, Aero/View)

• Airagon

• Western Aerial Contractors

• Park (Twinplex, unnamed)

Some, but not all, of these used a direction of flight indicator similar to the one shown below on P5. Some models also had early data strip styles, with both serial number and calibrated focal length included in the image (the example below, from a Park camera, has these on the left side of the frame).

Marker Separation (n = 120 reports)

markers	mean	median
P5 - P6	222.698 ± 0.489	222.49
P7 - P8	222.773 ± 0.659	222.495

flat wing (large)

A larger version of the “flat” wing was used in at least the Fairchild F-224 and K-17A cameras. In the lower-right corner of the frame, there is sometimes an indicator with the focal length (when not overwritten by other markings on the print, that is).

Note also the diamond-shaped marker on P5, which I am taking to be the “direction of flight marker” as the arrow-shaped notches on both P5 and P6 point in opposing directions.

Marker Separation (n = 1 reports)

markers	mean	median
P5 - P6	217.490 ± nan	217.49
P7 - P8	217.510 ± nan	217.51

flat wing (tall)

A “tall” version of the “flat” wing was used in primarily Aero Services (also labeled as Aero, Aero/View) cameras. Some versions of these cameras have a small marker that includes the camera number on it, usually just below P5. Note also the diamond-shaped marker on P5, which I am taking to be the “direction of flight marker”.

Marker Separation (n = 5 reports)

markers	mean	median
P5 - P6	222.596 ± 0.984	222.41
P7 - P8	223.010 ± 0.557	223.02

flat wing (curved)

This design was used primarily in cameras manufactured by Wes Smith (or just “Smith”). Note that marker P5 in the example below has a “direction of flight” indicator, meant to point in the direction that the plane was flying (in this case, towards the right of the frame):

Marker Separation (n = 6 reports)

markers	mean	median
P5 - P6	221.322 ± 1.038	221.51
P7 - P8	221.598 ± 1.423	222.055

curved wing

This design was used in at least some versions of the K-17 “modified” camera. Note that three of the markers have notches on either side, or are slightly inset from the rest of the frame - in the absence of any other identifying marks, I am taking P5 to be the odd one out.

Marker Separation (n = 4 reports)

markers	mean	median
P5 - P6	222.295 ± 0.135	222.345
P7 - P8	222.330 ± 0.076	222.355

sharp wing

This design was used in at least the Fairchild K-17 and some “modified” versions of the K-17 camera. As with the F-224 and K-17B cameras, there is a strip indicating the calibrated focal length in the lower right-hand corner of the frame, as well as a “direction of flight” marker on P5:

Marker Separation (n = 3 reports)

markers	mean	median
P5 - P6	226.160 ± 0.694	225.88
P7 - P8	226.063 ± 0.808	225.87

Fairchild Cameras

arrow-type fiducials

Caution

I have been able to identify a single example of this type of fiducial marker connected to a calibration report, and a number of others to notes made on the edges of the frame of an image from a separate survey, or to the film identification information taken at the beginning of a roll.

The reported camera in all cases was a Fairchild K-17, but it is unclear whether the K-17 also used other fiducial marker patterns (as appears to be the case based on calibration reports downloaded through Earth Explorer). As with the wing-type examples shown above, therefore, I recommend using estimate_measures_camera in the absence of any definitive marker coordinates.

These cameras have four mid-side fiducial markers. Both P5 and P6 are “arrows” that (typically) point in the direction of flight. P5 is the tip of the arrow-shaped notch that is level with the frame, while P6 is the tip of the arrow-shaped notch that is cut into the frame.

Many prints (or scans) of images of this type will cut off the point of the notch on P7 and P8, so it is probably better to use the point where the vertical edge of the notch intersects the image frame, rather than the “point” of the notch.

notch-type fiducials

mid-side only

This pattern was used in a number of different cameras, including T-11, T-12, KC-1, KC-1B, and CA-14. Note that the marker location is given by a small pinhole just inside the frame of the image, rather than the notch-shaped marks in the frame. Note also the distance between P5 and the principal point is longer than the distance between P6 and the principal point.

Marker Separation (n = 27 reports)

markers	mean	median
P5 - P6	237.946 ± 0.223	237.94
P7 - P8	235.035 ± 0.206	235.016

Marker Location (n = 2 reports)

name	x	y	angle
P5	0.000 ± 0.063	117.521 ± 0.007	179.995
P6	237.846 ± 0.050	117.521 ± 0.007	0.005
P7	120.428 ± 0.001	0.000 ± 0.038	89.995
P8	120.421 ± 0.007	235.000 ± 0.053	270.001

T-12

Similar to the pattern shown above, some versions of the T-12 also used corner fiducial markers. Note that the fiducial marker for each of these is a small pinhole just inside the frame, rather than the notch-shaped marks in the frame.

Note

In some of these models, P5 is further inside of the left-hand side of the frame, which is why the variability of the P5-P6 separation, and the x location of P5, is much higher than for the other markers.

Marker Separation (n = 5 reports)

markers	mean	median
P5 - P6	236.403 ± 1.657	235.201
P7 - P8	235.236 ± 0.137	235.205
P1 - P2	328.163 ± 0.850	328.256
P3 - P4	328.156 ± 0.838	328.24

Marker Location (n = 3 reports)

name	x	y	angle
P1	3.601 ± 0.335	233.519 ± 0.296	224.999
P2	235.453 ± 0.321	1.622 ± 0.303	45.012
P3	3.619 ± 0.302	1.664 ± 0.372	135.004
P4	235.517 ± 0.274	233.519 ± 0.296	315.007
P5	0.000 ± 1.694	117.645 ± 0.083	180.033
P6	237.211 ± 0.118	117.481 ± 0.079	0.046
P7	119.465 ± 0.072	0.000 ± 0.112	90.039
P8	119.613 ± 0.059	235.248 ± 0.113	270.033

The T-12 camera also used cross-type fiducial markers instead of dots, with the same approximate measurements:

T-5

Based on diagrams like this one, which are included in some calibration reports, I am taking the location of each fiducial marker to be the corner formed by the horizontal/vertical edge of the notch, and the image frame:

If the data strip is not visible on the image, you should be able to identify P5 as being a slightly larger than the other three notches.

Marker Separation (n = 4 reports)

markers	mean	median
P5 - P6	229.688 ± 0.219	229.645
P7 - P8	229.950 ± 0.014	229.955

KC-6A

The Fairchild KC-6A used four mid-side wing/notch-shaped fiducial markers. In addition to P7 and P8, there are two markers that are not included in the calibration report; these are off-center, and closer to P5 than to P6.

On most examples of the KC-6A that I have seen, there is a direction of flight arrow at the bottom of the area where the serial number and other ancillary information is shown (in the upper left of the example below).

It is not completely clear whether the marker is meant to be a small dot just inside the frame, as on the T-11 and T-12 examples above, or the tip of the notch cut into the frame. Based on the longer separation distance for P7-P8 compared to P5-P6 in the example calibration report, and the fact that in the example below, dots are visible on P7 and P8 but not on P5 and P6, I have used the dots for P7 and P8 but not for P5 and P6.

In other images, no dots are visible, though this may be due to issues with exposure/development of the film. Because of this, and the lack of scanned calibration reports for this type of camera, it might be better to use the tip of the notch, as is done for P5 and P6, and use estimate_measures_camera.

Note

There is a single calibration report for a Fairchild KC-6A in the USGS calibration report database, but it is mislabeled as a KC-8A.

Based on information taken from a film identification image (EarthExplorer ID: ARHHORNT110X172), the Fairchild F-489 “Terrain Objective Camera” (TROC) used a similar pattern, but did not have an identification strip with the camera make/model and serial number exposed. There are also no F-489 cameras available from the USGS database.

Marker Separation (n = 1 reports)

markers	mean	median
P5 - P6	234.893 ± nan	234.893
P7 - P8	236.089 ± nan	236.089

Marker Location (n = 1 reports)

name	x	y	angle
P5	0.000 ± nan	117.939 ± nan	180.018
P6	234.893 ± nan	117.939 ± nan	359.982
P7	117.147 ± nan	0.000 ± nan	90.014
P8	117.231 ± nan	236.089 ± nan	270.027

checker-type fiducials

T-11

Some versions of the T-11 used checker-type fiducials like the one shown below. Note the direction of flight indicator next to the P5 marker:

Park Cameras

In addition to the wing-style fiducial marker pattern, Park Aerial Surveys, Inc. manufactured and used cameras with eight crosshair-style fiducial markers (four corner, four mid-side).

These look very similar to the Wild Heerbrugg RC8 and RC10 crosshair-style cameras shown below, but are distinguished by the corner fiducial markers being closer to the frame (further from the principal point) and set inside of a smaller rounded frame:

Some, but not all, of these cameras also displayed either the camera serial number and focal length, or just the focal length, on the side of the frame near one of the fiducial markers.

Marker Separation (n = 12 reports)

markers	mean	median
P5 - P6	223.781 ± 0.039	223.786
P7 - P8	223.793 ± 0.032	223.802
P1 - P2	316.532 ± 0.033	316.537
P3 - P4	316.524 ± 0.028	316.52

Marker Location (n = 6 reports)

name	x	y	angle
P1	0.000 ± 0.021	223.825 ± 0.015	224.995
P2	223.808 ± 0.022	0.000 ± 0.016	45.009
P3	0.005 ± 0.009	0.028 ± 0.017	135.001
P4	223.821 ± 0.017	223.825 ± 0.015	315.004
P5	0.033 ± 0.018	111.922 ± 0.016	179.996
P6	223.804 ± 0.035	111.909 ± 0.014	0.011
P7	111.900 ± 0.019	0.037 ± 0.019	90.006
P8	111.932 ± 0.012	223.821 ± 0.020	270.01

Wild Heerbrugg Cameras

corner-only fiducials

Earlier Wild RC cameras, such as the RC5 or RC5A, or earlier versions of the RC8 or RC9, have four cross-shaped fiducial marks in the corners, set inside of a rounded frame:

Some versions also used an alternate corner marker:

Marker Separation (n = 48 reports)

markers	mean	median
P1 - P2	299.815 ± 0.011	299.813
P3 - P4	299.818 ± 0.013	299.818

Marker Location (n = 15 reports)

name	x	y	angle
P1	0.000 ± 0.010	212.002 ± 0.008	225.001
P2	211.999 ± 0.014	0.000 ± 0.015	45.000
P3	0.000 ± 0.012	0.001 ± 0.010	135.001
P4	212.006 ± 0.007	212.002 ± 0.008	315.000

Marker Separation (alternate corner, n = 2)

markers	mean	median
P1 - P2	299.842 ± 0.003	299.842
P3 - P4	299.798 ± 0.017	299.798

corner + mid-side fiducials

Later versions of the RC8 camera, as well as the RC10 camera, used a pattern with eight fiducial markers: four corner markers as on the corner type, and four mid-side markers.

These markers may be all larger crosses, as in the corner marker types (these were typically RC8 cameras):

Marker Separation (n = 32 reports)

markers	mean	median
P5 - P6	219.995 ± 0.008	219.996
P7 - P8	219.998 ± 0.009	220
P1 - P2	299.815 ± 0.009	299.813
P3 - P4	299.805 ± 0.008	299.803

Marker Location (n = 15 reports)

name	x	y	angle
P1	3.998 ± 0.010	215.993 ± 0.010	224.998
P2	216.004 ± 0.006	4.000 ± 0.009	44.999
P3	3.998 ± 0.006	3.997 ± 0.011	135
P4	215.989 ± 0.010	215.993 ± 0.010	314.999
P5	0.000 ± 0.006	109.996 ± 0.012	179.998
P6	219.996 ± 0.010	110.000 ± 0.009	359.999
P7	110.004 ± 0.006	0.000 ± 0.010	89.998
P8	109.994 ± 0.013	220.001 ± 0.012	269.997

Or, the markers might be a mix of cross-style markers in the corner, and crosshair style on the mid-side markers (again, these are typically RC8 cameras):

Marker Separation (n = 10 reports)

markers	mean	median
P5 - P6	220.000 ± 0.011	219.999
P7 - P8	220.005 ± 0.008	220.004
P1 - P2	299.815 ± 0.009	299.814
P3 - P4	299.810 ± 0.009	299.812

Marker Location (n = 9 reports)

name	x	y	angle
P1	4.002 ± 0.012	216.000 ± 0.009	224.999
P2	216.006 ± 0.013	4.000 ± 0.012	45
P3	4.002 ± 0.007	3.999 ± 0.009	134.999
P4	215.997 ± 0.009	216.000 ± 0.009	314.999
P5	0.000 ± 0.009	110.005 ± 0.015	180.002
P6	220.000 ± 0.012	109.999 ± 0.011	0.001
P7	110.006 ± 0.012	0.000 ± 0.012	89.998
P8	109.996 ± 0.009	220.004 ± 0.010	269.997

Or, they might be all crosshair style markers (typically RC10 cameras):

Marker Separation (n = 23 reports)

markers	mean	median
P5 - P6	220.008 ± 0.048	219.999
P7 - P8	219.998 ± 0.009	220
P1 - P2	299.814 ± 0.008	299.814
P3 - P4	299.807 ± 0.012	299.806

Marker Location (n = 16 reports)

name	x	y	angle
P1	3.999 ± 0.013	215.998 ± 0.006	225
P2	216.003 ± 0.014	4.002 ± 0.007	44.998
P3	4.005 ± 0.012	4.002 ± 0.006	134.997
P4	215.992 ± 0.013	215.998 ± 0.006	315
P5	0.000 ± 0.012	109.999 ± 0.011	179.999
P6	220.012 ± 0.060	109.998 ± 0.009	0.002
P7	110.002 ± 0.012	0.000 ± 0.007	89.996
P8	109.997 ± 0.013	219.996 ± 0.007	270.001

Zeiss RMK Cameras

Note

The model name for the Zeiss RMK includes information about the lens and film type. For example:

• Zeiss RMK 15/23 means that the camera uses a Pleogon lens with a ~15 cm (~150 mm) focal length and 23 cm film.

• Zeiss RMK A 15/23 means that the camera uses a Pleogon A lens with a ~15 cm focal length and 23 cm film.

• Zeiss RMK A 21/23 means that the camera uses a Pleogon A lens with a ~21 cm focal length and 23 cm film.

• Zeiss RMK A 30/23 means that the camera uses a Pleogon A lens with a ~30 cm focal length and 23 cm film

… and so on. This pattern doesn’t necessarily tell you the pattern of the fiducial markers, but it should at least give you a rough idea of the focal length and film size of the camera.

mid-side only

Earlier models tended to use only mid-side fiducial markers:

Marker Separation (n = 41 reports)

markers	mean	median
P5 - P6	225.995 ± 0.062	226.001
P7 - P8	226.008 ± 0.038	226.01

Marker Location (n = 5 reports)

name	x	y	angle
P5	0.000 ± 0.008	112.978 ± 0.009	180.002
P6	225.983 ± 0.016	112.978 ± 0.009	359.998
P7	112.990 ± 0.012	0.000 ± 0.021	90.004
P8	112.995 ± 0.009	225.973 ± 0.007	269.999

corner fiducial markers

Note

The separation distance and location for the corner fiducial markers (P1, P2, P3, and P4) for these cameras tends to be much more variable than the mid-side fiducial marker location (cf. \(\sigma > 0.3\) vs. \(\sigma < 0.05\)).

As with previous examples, it is likely a “safer” option to use estimate_measures_camera rather than the average measures below, at least for the corner fiducial markers.

Later(?) versions of the Zeiss RMK camera used eight fiducial markers: four mid-side markers, and four corner markers. These came in (at least) two main styles: a “floating style”:

Marker Separation (n = 60 reports)

markers	mean	median
P5 - P6	225.991 ± 0.024	225.992
P7 - P8	225.993 ± 0.028	225.995
P1 - P2	294.470 ± 0.735	294.101
P3 - P4	294.461 ± 0.713	294.112

Marker Location (n = 20 reports)

name	x	y	angle
P1	8.800 ± 0.368	217.162 ± 0.326	224.99
P2	217.161 ± 0.362	8.837 ± 0.332	45
P3	8.815 ± 0.343	8.797 ± 0.365	134.996
P4	217.136 ± 0.324	217.162 ± 0.326	314.992
P5	0.000 ± 0.015	112.993 ± 0.037	179.998
P6	225.981 ± 0.018	112.996 ± 0.029	0
P7	112.999 ± 0.033	0.000 ± 0.014	90.001
P8	112.998 ± 0.038	225.983 ± 0.023	269.998

and a “fixed” style:

Marker Separation (n = 41 reports)

markers	mean	median
P5 - P6	225.994 ± 0.016	225.995
P7 - P8	225.993 ± 0.011	225.994
P1 - P2	295.665 ± 4.064	294.018
P3 - P4	295.702 ± 4.062	294.027

Marker Location (n = 33 reports)

name	x	y	angle
P1	8.471 ± 1.590	217.527 ± 1.588	225.001
P2	217.521 ± 1.593	8.470 ± 1.592	45.001
P3	8.464 ± 1.590	8.473 ± 1.593	135.002
P4	217.528 ± 1.591	217.527 ± 1.588	315.001
P5	0.000 ± 0.010	112.998 ± 0.013	179.999
P6	225.992 ± 0.018	112.991 ± 0.012	0.004
P7	112.993 ± 0.013	0.000 ± 0.012	90.002
P8	113.002 ± 0.014	225.993 ± 0.009	270.003

Note

The coordinates for P5-P8 above correspond to the center of the small dot near the tip of the fiducial marker.

notes and references

[1]

Fleming EA (1960) Recognition of Air Survey Lens Types. The Canadian Surveyor 15(2), 91–96. doi: 10.1139/tcs-1960-0027.

[2] (1,2)

In the absence of any other identifying information such as a serial number on the frame, I am taking the camera information from the calibration report at face value. Because I have seen a number of instances where the camera information from the associated calibration report (make/model, serial number, lens serial number) did not match the information visible on the frame, however, the information about which cameras used which fiducial marker patterns, especially for the wing-type fiducial markers, may not be fully accurate.

[3]

In a number of the calibration reports for these “modified” cameras, the phrase “… which has been modified to meet the requirements of a precision camera” is included. Presumably, this modification is adding fiducial markers, which might explain why this camera type corresponds to so many different fiducial marker patterns.