Requesting for new feature

[RollaHamza]

Greeting, I wonder if it possible to extract data from map figure

[nbehrnd]

You do not clarify what kind of data you want to extract. Since it is a map with discrete color levels (as antonym to continuous), perhaps you are interested to count the countries associated to gray vs. countries associated with different tints of blue. In terms of resolution, you might run into troubles: Apparently (e.g., Egypt, Libya, Chad, Sudan) share one level of a property and still are discernible as individual countries by a gray line marking their frontiers. Possibly, this line might be too thick to allow the recognition of countries of smaller surface (e.g., Slovenia, Croatia, Kuwait, Singapore, Maldives to mention a few).

In case you were interested in the area; an implementation of a planimetry-like approach has been suggested earlier (here and here). At present, I speculate this is a task perhaps better resolved with ImageMagick (note there is a dedicated tag for questions about this program e.g., on stackoverflow). However -- a big however -- one would need to correct for the distortion the map introduces when displaying a 3D object in 2D.

[RollaHamza]

I want to know for example which countries is shaded in certain color another example is below i want to know the name of countries that is in red color

[nbehrnd]

Multiple options are available but (likely) neither one is in the scope of the digitizer's design.

• the map with tints of blue looks like generated by R/ggplot2. Identify the corresponding author(s) of the illustration, justify your interest to access the underlying .csv data. In case of a scientific publication, authors often share these in the supplementary material of a publication on the servers of the publishers, a public repository (like zenodo, or figshare), or their institution.

• ImageMagick may substitute the color of pixels of one specific color by black (an example), so if the individual tints used in the chloropleth map differ enough in their hue, one may extract the individual levels. One could then use these «stencils» to attribute the countries manually by visual comparison with maps about larger regions/continents. It is achievable to recognize them e.g. with anki decks about geography like this one.

  In case there are many chloropleth maps, one may use the data from wikipedia e.g., the admin 0 maps (a brief how to) to 1) generate illustrations for training, i.e.: same projection, orientation and scale as in the maps you want to analyze where the shape of one country after an other is filled black (like in the illustrations to analyze, but already processed by ImageMagick). Then 2), one would train a computer program with the data derived from the wikipedia data. (I'm not versed in this topic, but Python/opencv is one of the environments reportedly used with success for pattern and face recognition.) Eventually 3), the trained computer program could be able to recover the assignments ("0.8 units: Poland, Slovakia, Czech Republic ; 0.2 units: Slovenia, Croatia", etc) at faster rate than spot by eye. Obviously, this third option requires the most initial investment.