Yahya, what is "better precisity" in georeferencing? I mean, the methods and the tools are more or less the same; in my experience, accuracy depends on your own accuracy and 'a priori' knowledge (projection, datum, sheet system) rather than the specific software used.

ERDAS is better that ArcGIS in one specific aspect: it enables to define user datum (which seems to be impossible in ESRI) and the menu system of the new coordsys definition is quite easy to understand and to use. However, from this point of view, Global Mapper is also on this level of easy usage - and it is much more cheaper.

I would suggest the Global Mapper. It is not free, however quite cheap, easy to use, easy to define user grids and datums and knows a lot of output formats. If you are familiar with the GDAL terminology and coordinate system codes, the QGIS (Quantum GIS) could also be suggested from the open source set.

I thought the global mapper was highly useful in file conversions only. Never used it for the georeferencing purpose. Does it have an extensive set of datums ( i.e more than that defined in Erdas Imagine??)

ERDAS Imagine - Geometric correction for Image to image registration , ERDAS Leica Photogrammetry Suite for Orthorectification is good. ArcGIS we can use for topo registration purpose.

I appreciate and agree with Srijith Balan, no doubt Erdas is a better option, but if you can use ArcGIS efficiently and know the image's datum and ellipsoid used to reduce to a specific datum, preferably use ArcGIS.

Dear colleagues, the coordinate system dataset and the georeferencing capability of the softwares are two, very different pieces of cake.

If you look for datums and/or coordinate systems, starts from this:

http://www.epsg.org/geodetic.html

An important question is that how easy can you define a new CS or datum in your GIS. In QGIS or Global Mapper, it is pretty easy, however defining a new datum in ArcGIS is a nightmare (or at least it was in the series 9, I haven't tried it later).

Sir,

can you kindly elaborate

To follow of on the last post, PostGIS is a good open source option. In my ex perience, ArcGis probably provides the largest number of options (datums, projections, and coordinate systems). In addition, ArcGis gives you the flexibilty to define custom projections. Hope this helpful

I completely agree with my friend Gabor. You can also look at Map Window that report completely the parameters estimated for the georefencing but has less options (so it seems) than QGIS for example. Be carefull with predined datum transformations in commercial packages: I observed very unaccurate transformations and sometimes some misinterpretations of the actual projectionand datum: be careful!

I work well with Mapinfo. Simple to use, interoperability with many other systems, databases, web etc.. It does not cost much, and it is easy to learn.

Obviously ArcGIS is great, but we are on a different level

However, if anybody knows how to define a new datum in ArcGIS, I'd be extremely happy with the answer :-)

You will have to use an existing datum available in ArcGis..but you can manipulate the map projection, points of tangency, central meridians, units, and coordinate systems. You can do this using Project Define, or Project, both in the ArcGIS toolbox

Thanks Mark! According to my best knowledge, it is indeed not possible to add a new user datum to the ArcGIS - which is a clear barrier to me, that's why I suggest other softwares for *georeferencing*

It depends on what kinf of data you have. Certainly, Erdas is better if you want to georeference satellite images. Also, ENVI is very good, though.

Depends by your degree of experiences. ArcGis is the best but also Global Mapper is good.

I prefer Global Mapper to georeferenced, but I especially georeferenced maps feature in projection system specific Romanian Maps (Stereographic, Datum S-42, or Dealul Piscului EPSG 31700) and works very well. It's intuitive and easy to operate. You can create your own datum if necessary or can enter an EPSG code to be recognized by the program.

I prefer ARC GIS but it depends of your experience. Autocad it is a good and easy tool to georeferece maps before edits.

For simple or flat area I use Global Mapper. For hilly terrain I use ArcGIS

ArcGIS is the perfect tool...

There is a lot of GIS software for georeferencing like ArcGIS, Geomedia, MapInfo etc but I suggest Global Mapper software (http://www.bluemarblegeo.com/products/global-mapper.php). GM is very cheap and have a lot of possibilities.

In response to John Acubia - ArcGIS from the point of view georeferencing purposes is not a perfect tool. Georeferencing tools in ArcGIS are probably the weakest modules in ESRI software.

I completely agree with the the answer given by Valerio Baiocchi. Beyond software and algorithms you should carefully check the datum parameters used for transformations both in commercial and free software. It saves you terrific amounts of work and time if you need to overlay different data sources particularly historical maps.

from my experience, I found ERDAS Imagine is better than ArcGIS and more precise.

Yahya, what is "better precisity" in georeferencing? I mean, the methods and the tools are more or less the same; in my experience, accuracy depends on your own accuracy and 'a priori' knowledge (projection, datum, sheet system) rather than the specific software used.

ERDAS is better that ArcGIS in one specific aspect: it enables to define user datum (which seems to be impossible in ESRI) and the menu system of the new coordsys definition is quite easy to understand and to use. However, from this point of view, Global Mapper is also on this level of easy usage - and it is much more cheaper.

Still completely agree with Gabor and also with Fabio. Moreover You have to keep in mind the difference between accuracy and precision: in georeferencing precision is how much your points are coherent with themselves while accuracy is how much You're near the correct value. The most significant parameter is obviously the accuracy because it hints the error You 'll have using the georeferenced cartography. You cannot evaluate precision with programs that allow you to estimate the 'error' only on points that You used to georeference (in photogrammetry they are called Ground control points) because you underestimate Your 'error'. You need a software that allow You to evaluate 'error' on points not used to georeference (in photogrammetry they are called Check points) . ArcGIS hasn't CPs, Qgis has CPs so its prevision of the final 'error' is not underestmated. Sorry if I was a little "professorial" but I can be different: I'm a professor! ;-)

Thank you for all your updates ,

and Mr. Valerio ,even though i'm familiar with GCP, the concept of check point is new to me and thank you for introducing that to me. The feedback provided by all of you have been immensely helpful and I believe that it will continue.. :)

Not at all Srijith. Here is one of my more recent paper on the subject, I'm planning to write something of more comprehensive.

http://technicalgeography.org/index.php/latest-issue-1-2013/68-02_valerio_baiocchi_keti_lelo_maria_vittoria_milone_martina_mormile

I suggest Global Mapper which is easy, quick and simple.

Yes, I repeat, my personal favourite is also Global Mapper. However, the new (?) Georeferencer module of Quantum GIS is very appealing and effective, if you're familiar to GDAL formalism of projections - and it is free.

What I have meant is that ERDAS is more flexible, easier and predifined the parameters (which you can define the datum, the geometric method for transformation and you can see the error in every point).