How to get image position (X, Y)

[Michael Fliorko]

Hello everyone,

I have pdf-file with text and images. Let say products' descriptions and photos.
Some photos consist of several adjacent images.

I want to extract products' photos and for each product join all photo parts into one image.
At the beginning I need to know which images are adjacent.

I studied sample source code and see that I can read image size (height, width). But I haven't found how to read image position on the page.

Could someone give me a clue how to get image positions from existing PDF file?

Thanks,
Michael

[Michael Fliorko]

It seems I found a solution for ImageXObject in the sample ContentScanningSample:

          if(xObject is xObjects::ImageXObject)
          {
            Console.WriteLine(
              "Image '" + xObjectKey + "' (" + xObject.BaseObject + ") " // Image key and indirect reference.
                + "on page " + (page.Index + 1) + " (" + page.BaseObject + ")" // Page index and indirect reference.
                );

            // Get the coordinates of the image!
            double[] ctm = level.State.CTM; // Current transformation matrix.
            SizeF imageSize = xObject.Size; // Image native size.
            Console.WriteLine("  Coordinates:");
            Console.WriteLine("     x: " + Math.Round(ctm[4]));
            Console.WriteLine("     y: " + Math.Round(page.Size.Value.Height - ctm[5]));
            Console.WriteLine("     width: " + Math.Round(ctm[0]) + " (native: " + Math.Round(imageSize.Width) + ")");
            Console.WriteLine("     height: " + Math.Round(Math.Abs(ctm[3])) + " (native: " + Math.Round(imageSize.Height) + ")");
          }
        }