I posted recently about confusion in the geospatial community around the terms DSM, DEM, and DTM. Another set of terms that are used in different ways by different people are those that refer to the geo information associated with an image. I thought it might be useful to outline the way that I use these terms. Be aware that the ENVI documentation, and other Exelis VIS employees, as well as plenty of people in the geospatial community, sometimes use these terms differently. But if I were king of the world, I’d define them like this:
Georectify: To take an image that has not been adjusted to be in a known coordinate system, and put it into a known coordinate system. Usually this means taking an image that is in its original geometry, and putting it into a map projection. There are different ways to do this. Perhaps the most common way is to identify a set of points in the image for which the latitude and longitude or map coordinates are known, and use them to warp the image into a map projection.
Georeference: To take an image that is already in a known coordinate system, and provide the information necessary for software to understand which coordinate system it is in.
Geocode: Same as georeference.
Geometrically Correct: Same as georectify.
Orthorectify: To take an image in its original geometry and very accurately adjust it so that it is in a known coordinate system, with distortions due to topographic variation corrected. An orthorectified image has uniform scale throughout the image. A DEM (and by this, I mean an image in which the pixel values represent the ground elevation above sea level) is required for true orthorectification.
How do you use the terms georeference, georectify, geocode, geometrically correct, and orthorectify?
ImagerySpeaks 
Peg, I really found this post and your post on DEM vs DTM vs DSM very useful. I’d love to see you keep this theme going as there are many instances where people are talking past each other because of different terminology.
Thanks Brian. I appreciate hearing that it’s helpful. I’ll think about where else we run into terminology confusion, and see if I can write about those as well. Does anyone have any suggestions about other terms that sometimes get confused?
“Scientific visualization” comes to mind. 😉
Thanks for the article. I was just researching the difference between georeference and georectify last week and was amazed at the number of times I saw georeference and georectify listed as being the same! Could you clarify georeference where you mention “provide the information necessary for software to understand which coordinate system it is in”? In the case of ArcMap, is this just a matter of making sure some type of projection file is present or using the georeferencing toolbar or using a reproject tool?
A suggestion for terms of confusion would be projection vs coordinate system.
Hi Charlotte,
Yes, I’m always surprised when people say they are the same, too! In terms of how to provide the information necessary for software to understand the coordinate system, the way the info needs to be provided will be different for different software packages. For ENVI it is done by entering appropriate information (e.g., which projection, which datum, what units, pixel size, and the coordinates in that projection of a pixel in the image) into the header fields that are intended to hold that information. You can either manually edit the header in a text editor, or you can bring up ENVI’s Header Info dialog (by right clicking on the file in the Available Bands List, then choosing Edit Header), then choose Edit Attributes > Map Info. You’ll then get a dialog that allows you to specify all of the necessary information.
I am not as familiar with the ESRI products, but I believe that if you have a dataset that is in a projection, but doesn’t have the projection information contained anywhere in the metadata (i.e., it needs to be georeferenced), you can add that information in ArcCatalog. It looks like what you do is select the dataset in the Catalog Tree, then right click and choose Properties. Scroll down until you see the Spatial Reference section, which should be shown as . Click the Edit button on the right of that line, and you will be given a dialog from which you can set up the coordinate system info. ESRI can probably give you more details about how this works in their products (or correct me if I’m wrong).
I agree that the way ESRI uses the terms “coordinate system” and “projection” can be confusing. In the ENVI world, we don’t really talk about “coordinate systems”. Everything in ENVI is called a “projection”. Although, technically, the Geographic Lat/Long “projection” in ENVI is not really a projection at all, in that it is a coordinate system defined for a sphere, not for a 2 dimensional representation. ESRI makes the technically correct distinction between coordinate systems for a sphere, and projected coordinate systems.
I hope that helps!
Peg
I consistently misuse these terms, so I appreciate this explanation.
Please, allow me to disagree when you say “Geocode: Same as georeference”.
Geocoding can be broadly defined as the assignment of a code (usually geographic coordinates: latitude, longitude) to a location, usign other source of geographic data.
Georeferencing is to establish a location in terms of map projections or coordinate systems. It involves aligning geographic data to a known coordinate system so it can be analyzed, viewed, and queried with other geographic data.
So, the main difference is that in the georeferencing process there’s, usually, a map projection. For example, an image that is not projected to a map, with (lat, lon) info in each pixel, is geocoded. The geographic data can be transformed and augmented according to a given coordinate system.
Hi cmdelatorre. Thanks for your comment. I like how you make a distinction between images that have location information, but are not actually *in* the orientation of the known coordinate system. So, if I understand you correctly, images that have RPC data (so in ENVI you can see lat/long or map coordinates for any point in the image) would be geocoded, but not georeferenced. That makes sense to me.
So, OK, now if I were king (actually, queen) of the world, I’d add that distinction to my definitions.
Thank you!
Peg
Are all these used in geometric correction of an image?
That’s another term that people use in various ways. Some people use it when they’re adjusting an image’s orientation to match another image, even it that image isn’t in a map projection. It is often used to mean that one is putting an image into a map projection. I can imagine that someone could also say they are applying a geometric correction if they were taking an inaccurately georectified image and improving the georectification (for example, by orthorectifying it). ENVI 5 uses the term in a very general way, grouping all of the tools that adjust an image’s spatial orientation together into a Geometric Correction category.
Hello Sir,
While I will return to the differentiation of the terms in this article, I am in need of understanding the difference between “Terrain corrected image Vs Orthorectified Image”.
Thanking You in advance,
Sekhar
I have mainly hear the term “terrain corrected” applied to SAR imagery, which has distortions caused by topography and the side-looking geometry of the sensors. So, in this sense, terrain corrected would mean that it is corrected for effects like layover. I also believe that GeoEye has used the term terrain corrected when talking about orthorectified products. So, sometimes the two terms are used interchangeably. Have any of our readers heard other usage of the term “terrain corrected?”
Hi,
I have two files. One is a image file (QB) which is geo-referenced (i.e., it has known coordinate system) and a vector file (eg., road file) also in known coordinate system of the same area. When i overlay these two files on each other in ArcMap, they are offset slightly and I am unable to decide which one of them to use to correct the other. Can anyone comment?
Thanks
Meghan
Hi Meghan,
It really depends on which has the more correct georeferencing, and it’s hard to tell from what you’ve said so far. Do you know anything about where the geolocation information for either dataset came from? In other words, how was the vector file made, and how was the image georectified? Do you have any very reliable independent maps etc. of the area, to which you can compare both datasets?
Peg
Thanks Peg. I dont have any other maps unless i can use google maps. Both the datasets are from reliable sources.
Hi Peg,
check out the ArcGIS Dictionary. They also distinguish between each term in a comprehensible way.
http://support.esri.com/en/knowledgebase/GISDictionary/term/georeferencing
http://support.esri.com/en/knowledgebase/GISDictionary/term/georectification
http://support.esri.com/en/knowledgebase/GISDictionary/term/orthorectification
In your opinion which step is the first to be applied to raw satellite data, georeferencing or georectification? Can you explain why?
Thanks,
Peter
Hi Peter,
I would say that according to my definitions, georectification usually includes georeferencing. Because otherwise you’d be putting the data into a projection (georectification), but not saving any metadata about what that projection is (georeferencing). Which wouldn’t make much sense.
But the ArcGIS Dictionary defines georeferencing in a way that includes actually adjusting the orientation of the data. I use the term differently.
Peg