Use sumatrapdf-tool.exe or SumatraPDF.exe with the same arguments.
Available in pre-release 3.7
You can use SumatraPDF from the command line to extract images embedded in a PDF file.
To extract all images: sumatrapdf-tool extract input.pdf
It will save each image in a file named image-<object-number>.png.
The image extension depends on the file type; for example, JPEG images will be saved as .jpg files.

To extract specific images #

You can extract only specific images. In a PDF file, each image has an object number.
First, you need to find the object numbers of the embedded images: sumatrapdf-tool info -I input.pdf
The output looks like:
Images (9):
        1       (1189 0 R):     [ Flate ] 400x300 8bpc DevRGB (3 0 R)
        1       (1189 0 R):     [ Flate ] 400x300 8bpc DevRGB (6 0 R)
        1       (1189 0 R):     [ Flate ] 602x250 8bpc DevRGB (10 0 R)
        1       (1189 0 R):     [ Flate ] 278x288 8bpc DevRGB (15 0 R)
3, 6, 10, and 15 are object numbers.
To extract images 3 and 15: sumatrapdf-tool extract input.pdf 3 15
Reference: