Modern hard drives are typically rated for a million hours MTBF (mean time before failure), and they'll do that, provided you look after them (no moving them around, keeping a fan on them so they don't cook, no huge hot/cold temperature swings etc).
They're also good for linear reading and writing at up to about 90-100MB/sec; random reads/writes you're looking at about 100 seeks a second - thus, not so good if you're dealing with a ton of small files. CCTV is almost always linear reads/writes of huge amounts of data, so often times regular 7200rpm 1-2TB hard drives are absolutely fine - providing they're in well conditioned environments.
SSDs are insane; read more about how and why at the following link:
http://www.codinghorror.com/blog/2011/05/the-hot-crazy-solid-state-drive-scale.html
In short, they're usually at least twice as fast for reading and writing in linear workloads, and are about 100 times faster when dealing with tons of tiny files - as they can typically manage several tens of thousands of seeks per second. This is a big deal in applications like heavily used databases, caches, and thumbnails.
SSDs are a LOT more robust when it comes to temperatures, and being knocked about - HOWEVER there is a significant gotcha, which is their limited lifespan. If you run an SSD at it's rated maximum read/write rate, you'll probably get 9 months to a year out of it before it randomly dies - taking all your data with it. You can read more about this in the article linked above. The MTBF on SSDs, unfortunately, do not take into account this 'continuous read/write' consideration - it's usually on the overall assembily with either no or light usage. Read the specsheet on how they measured it to be sure.
None of this takes into account RAID, which if you don't want to lose data is a MUST. Raid 1 - aka 'mirroring' - should be an absolute requirement for you if you can't afford to lose your CCTV data; even MORE so if you're using an SSD, due to their aforementioned tendency to throw their toys out of the pram when they've had enough.
In your case, you're probably best going with two 1TB or 2TB 7200rpm drives from a manufacturer you trust (I use Western Digital's Black range, but like all decisions, it's a personal preference these days), and stick them in raid 1 and configure your RAID subsystem to email you when a drive dies.
Oh, one more thing - an 'enterprise' hard drive is just one which has had it's label changed, it's seek timeout (how long it'll click-click-click and retry trying to read data) dropped from 60 to 10 seconds, and a price tag of two to three times the regular price stuck on it.
One place I work at has over 400 western digital drives in use, and you'll always get a 2-5% failure rate on all hard drives in the first year, regardless of manufacturer. Our SSD failure rate in the first year is under 1%; however on our busier customer's servers it's almost 25% after the first year. We expect to lose another 30-40% of them in the next year on the busy machines that are thrashing the hell out of their SSDS - so again, unless you're dealing with tough environmental conditions, I still recommend regular hard drives in RAID.
let me know if you're looking for specific recommendations or advice; I build and maintain servers in a datacenter, so I tend to get a lot of exposure to when things go wrong.
