This was three years ago. If I was to do it over again, I would use Scrivener, since it has exporting to ebooks built in.
Once I had the framework, it was just a matter of producing the single images that went on each page. Now things can get more complicated by creating the text separately in the HTML, allowing for zoomable regions, etc., but I made my text clear enough that I didn’t require that. I chose to keep it simple for my first book. Perhaps I’ll get adventurous in my next picture book (yes, there’s another picture book waiting in the wings 😉 ).
The next step was to run all my files through the Kindle previewer, which converted my HTML and images into a .mobi file, which it then previewed on a simulated Kindle device. After I previewed my book, and made sure there were no errors, I uploaded the .mobi file to the KDP website. I could have also uploaded my zipped HTML/image files and word files and let Amazon do the conversion, but preferred to upload the .mobi file that the preview generated. It was uploaded exactly as is. I’ll elaborate on that process in more detail in part two.