I used LibreOffice to open Word doc and to save as HTML.
Till LibreOffice 4.1, the images are extracted and stored separately along with HTML file.
But, after LibreOffice 4.2, they moved to base64 type of encoding of images, so that images are embedded into HTML files. We can not separate images from HTML files.
This was so annoying and many people are reporting this as a bug here.
But, This seems not to be fixed.
So, I installed LibreOffice 4.1 in /opt just to use the old feature of storing images separately.
Just now, found another utility, unrtf to do the same.
To install it in ubuntu/debian
sudo apt-get install unrtf
If you get an word doc with images, save it as a Rich Text File using libreoffice writer.
test.doc -> test.rtf
unrtf test.rtf > test.html
This gives a nice HTML file and images separately.
Thanks for the GNU team for the nice utility.
I can get rid of old LibreOffice 4.1 now.