Realistic market projections depend a lot on matching your circumstances to the statistics being released.
(1) Projected growth for surveillance in North America is almost certainly under 10%. IMS recently ran a press release saying 12% but that's global and is skewed by China / developing countries.
(2) If you are looking at North America, the market is not split 50/50 in IP vs analog sales. IP is clearly significantly higher than analog.
Related, and a serious question, how would knowing how many units were being sold impact your decision making process?
(3) The number of integrator question is really fascinating. One immediate challenge you get into is who counts as an integrator? Let's you are talking about real companies with more than a few employees. It's certainly not 50/50 on sales. I think it's more like 80 / 20 in terms of sales. There's lot of integrators doing close to 0 analog sales today (outside of maintaining existing legacy systems).
(4) The market size for DIY systems is hard to estimate but here are some data points:
Those are not huge dollar numbers (Mobotix and Avigilon did ~$100 million in 2013, Axis did ~$750 million, etc.). However, given their far lower average selling price, that does indicate they are moving a lot of units.
We are doing a new survey round and one clear theme we see is that DIY solutions (whether its Costco, online kits, local tech shops, etc.) are becoming an increasing competitive problem.
Returning to your original question about North American growth. Overall, for surveillance, it is probably in the 6% to 9% range but once you factor in that a lot of sales are being diverted to the DIY channel, it's pretty clear that sales for traditional surveillance integrators, as a whole, are growing very slowly, if not flat.