Georeferencing points misplaced after import from Link Table

421
11
05-25-2020 03:03 AM
GeorgHohberg
New Contributor

I am using ArcGIS 10.3. A collegue in Asia georefferenced a .tif image for me using the ArcGIS georeferencing tool. He uses the same version of ArcGIS. Before rectifying he saved the link table. He send both, the link table and the georefferenced image. Trying to adjust some of the georeferencing, I opened the original image and imported the link table. After having imported the link table, the source points are about 2 km west and 2 km north of what they should be. I cannot figure out, why. My collegue and me both use the same coordinate system (WKID 32642, WGS_1984_UTM_zone42M). Does anybody know what could be the problem?

Thanks,

Georg

0 Kudos
11 Replies
DanPatterson
MVP Esteemed Contributor

not a feet vs meters thing is it?


... sort of retired...
0 Kudos
GeorgHohberg
New Contributor

Hey Dan,

as far as I know map units are determined by the coordinate system selected so I would not be able to change these. Changing the display units obviously does not help or do you have something else in mind regarding the units?

0 Kudos
DanPatterson
MVP Esteemed Contributor

The link table may have been made in the coordinate system of the map and not the data, hence, they could be different.  Worth a check on the values of the link table and how the coordinates are defined and whether they are correct


... sort of retired...
0 Kudos
GeorgHohberg
New Contributor

Hey Dan,

thanks for your efford. I checked the .prj of the Link table and the data frame properties. Both are identical.

Data frame property

.prj of Link Table

WGS_1984_UTM_zone_42N

PROJCS["WGS_1984_UTM_zone_42N"

WKID: 32642 Authority: EPSG

AUTHORITY["EPSG",32642]]

Projection: Transverse_Mercator

PROJECTION["Transverse_Mercator"]

false_easting: 500000,0

PARAMETER["false_easting",500000.0]

false_northing: 0,0

PARAMETER["false_northing",0.0]

central_meridian: 69,0

PARAMETER["central_meridian",69.0]

scale_factor: 0,9996

PARAMETER["scale_factor",0.9996]

latitude_of_origin: 0,0

PARAMETER["latitude_of_origin",0.0]

Linear Unit: Meter (1,0)

UNIT["Meter",1.0]

 

 

Geographic Coordinate System: GCS_WGS_1984

GEOGCS["GCS_WGS_1984"

Angular Unit: Degree (0,0174532925199433)

UNIT["Degree",0.0174532925199433]]

Prime Meridian: Greenwich (0,0)

PRIMEM["Greenwich",0.0]

Datum: D_WGS_1984

DATUM["D_WGS_1984"

Spheroid: WGS_1984

    Semimajor Axis: 6378137,0

    Semiminor Axis: 6356752,314245179

    Inverse Flattening: 298,25722356

SPHEROID["WGS_1984",6378137.0,298.257223563]]

Unfortunatelly, this does not seem to be the problem.

0 Kudos
DanPatterson
MVP Esteemed Contributor

What are the actual values of a location and for what location?  2 km is a big difference.  200m would suggest a datum difference,   What other coordinate systems are used in your area?  I have run out of ideas... what does your associate say?


... sort of retired...
0 Kudos
GeorgHohberg
New Contributor

These are some of the values from the link table. In Excell I calculated DX and DY. I find it noticable that the differences between X Source and X Map aswell as Y Source and Y Map are similar in magnitude to the error in the positioning of the map.

X SourceY SourceX MapY MapDXDY
442933,55384352762,69446045,7584350422,33112,203862340,3818
449572,9074350765,95452114,654348543,762541,742962222,19235
445476,24284345961,33448274,2154343946,742797,972382014,59736
447174,62064345953,34449935,144343867,242760,519792086,09858
446277,43014346276,96449007,474344264,312730,039652012,65578
442093,93174351992,88445288,3864349750,923194,454592241,95361
445759,62464341989,77448716,074340025,482956,445721964,291
444031,87154348378,63447038,9664346291,023007,094242087,6092
446004,97664349264,23448906,5714347095,162901,594022169,07068
444465,2864352336,61447527,4274349949,363062,141222387,25015
449300,09614341699,96452029,0174339853,692728,92131846,26766
445720,27174352941,05448695,2614350577,282974,989072363,77764
448935,27594349837,75451499,854347677,442564,573822160,30782
0 Kudos
DanPatterson
MVP Esteemed Contributor

I have no more ideas, except to revisit the georeferencing.  Sorry, the differences are too big to be something simple


... sort of retired...
0 Kudos
GeorgHohberg
New Contributor

Hey Dan,

after you had run out of ideas, I did no longer pursue this issue until today my collegue send a new link table which he used for georefencing a similar image. It looks like that:

9.543703    2.445142    444393.037816    4340604.876407
10.210346    4.964898    445233.124652    4343248.168412
11.711133    3.523411    446550.556075    4341673.463281
2.363828    7.071456    436840.989781    4345520.312199
10.966851    4.219995    445830.961943    4342445.428037

I can easiely recognize the map coordinates as the correct UTM coordinates. Yet, I do not understand what format the source coordinates are in. Source, target and data frame are set to WGS_1984_UTM_Zone_42N (WKID: 32642 Authority: EPSG). When I set source georeferencing points in the same project as my collegue, I get UTM coordinates. Do you or anybody else have any idea what could have let to the diverging coordinates system that the source coordinates are recorded in in my collegue's project? I am pretty sure that this is related to the original question of this thread.

P.S.: This is what the .prj file says:

PROJCS["WGS_1984_UTM_Zone_42N",GEOGCS["GCS_WGS_1984",DATUM["D_WGS_1984",SPHEROID["WGS_1984",6378137.0,298.257223563]],PRIMEM["Greenwich",0.0],UNIT["Degree",0.0174532925199433]],PROJECTION["Transverse_Mercator"],PARAMETER["false_easting",500000.0],PARAMETER["false_northing",0.0],PARAMETER["central_meridian",69.0],PARAMETER["scale_factor",0.9996],PARAMETER["latitude_of_origin",0.0],UNIT["Meter",1.0],AUTHORITY["EPSG",32642]]

0 Kudos
DanPatterson
MVP Esteemed Contributor

the first two columns look like image coordinates unless you are referencing images near the equator and the Greenwich meridian


... sort of retired...
0 Kudos