Here's the thing—continuous improvement and innovation are two sides of the same coin. The divide mostly comes from how we set up and separate work in organisations.
That doesn't mean there are no differences between the two. In my experience, meaningful distinctions are the risk profile, scope, and type* of idea.
The riskier and more uncertain the idea is, the bigger the scope is, and further from the core it is, more likely it is to benefit from the innovation process.
Ideas that focus on tweaking existing stuff (e.g. features and procedures), that take little time and resources, and are close to the core, are more likely to benefit from the continuous improvement process.
And don't confuse methods for either of above two.
Just because you are doing A/B testing doesn't mean you are innovating. If you are testing the colour of an existing log in button, then you are doing continuous improvement; if you are testing a landing page for a product you never sold before, you might be innovating.
At the end of the day, both continuous improvement and innovation are legitimate pathways to competitive advantage. Why not pursue them concurrently?
*Borrowing the terminology from Tuff and Nagji, I differentiate between core, adjacent, and transformational ideas.