![](https://secure.gravatar.com/avatar/ad7c76b9aa8758315397d0dfb15fc222.jpg?s=120&d=mm&r=g)
On Apr 16, 2012, at 9:59 PM, Pranav Lal wrote:
Many thanks for your responses. I am on gmail's web interface as of now so am responding to all the posts so far in a single message.
I am trying to imbed Images in an epub book. This book could be read on a variety of devices so I am hoping to use SVG to render the image correctly (for want of a better word) on each device the book is opened. The book will be transmitted over the Internet in a low bandwidth situation so I am trying to shrink the SVG files.
As for the conversion, I am using the -l commandline option to produce plane svg.
I would love to be able to convert some of the SVG to objects but have not found an automated tool to do this. I could write my own using shape recognition algorithms but that is my last resort.
That tells us on how you save the SVG once it had been created, but not how the SVG was originally done.
The usual way is to use the 'Trace Bitmap' functionality on an imported jpg and then removing the reference to the jpg. http://tavmjong.free.fr/INKSCAPE/MANUAL/html/Trace.html
The main place to gain size reductions is in the initial conversion. Sometimes this also can be aided by cleanup or simplification of the image before tracing.