[Turtle91]

I use the same basic procedure as BeckyEbook. However, I use the ReworkChapterHeads plug-in. That will automatically insert incremental IDs into a header tag.

Then, run a regex to incorporate the image into the header:

With:
<h3>1 (2, 3, 4, etc)</h3>
<p class="textbreak"><img src="../images/00003.jpeg" alt="" /></p>


Find: <h3>(\d+)</h3>\s*<p class="textbreak"><img src="../images/00003.jpeg" alt="" /></p>

Replace: <h3 title="\1"><img src="../images/00003.jpeg" alt="" /></h3>

I’d even rename the image file to something I could understand just by reading it eg. img_ChHd.jpg or something.

If your reading device is at all modern, and supports CSS, then you could even trim that down with:

Replace: <h3 title="\1"></h3>

and add css to incorporate the image as a background of the header tag; or even insert the image before the header tag. That is certainly not necessary, but it cleans up your code a little bit.

For example:

h3 {
display: block; 
background-image: url("../Images/img_ChHd.jpg"); 
background-size:2em 1em; 
width:2em; 
height:1em; 
margin:1em auto; 
content:""}

* set sizes as needed for the image

EDIT: added css example code