Segment Filtering

Providing cadence data wouldn't help you to distinct cycling and running, both have approximately the same value range.

For the pace there is already some (unreliable) system in place where new segment efforts with too fast speeds are automatically flagged.

minimum requirements to have Cadence and HeartRate.
Filter Out all unnecessary people.

Not many Cycling using Cadence

Just want to reduce Normal People.

normal People usually not fully equiped, and will not run fast like that.

so with the HeartRate & Cadence, we assume they are qualified to challenge Segments

@Jore928 - Really, only a fraction of even the faster people use all the things like heart rate or cadence.  I only recently started using a heart rate monitor and still often forget it or just don't bother when I go out, yet I have a lot of KOMs and CRs.  Are those now not valid since I didn't use a heart rate monitor?  On top of that, many people just use watches that only record wrist-based heart rate which is widely known to be very inaccurate much of the time.  You would never be able to tell if the posted heart rate was anywhere near reflective of the effort put in.  
The best thing is to look at the times.  If they are unrealistic, just flag them and they will be removed.  It's pretty easy to do.  Strava does have some automatic flagging, but they have a long way to go to improve it.  

I have Flag 1 lady many many times, because she always doing the same exercise everyday, beat the same Segment everyday, how do I flag her everyday...???

I can usually tell whether it was a bike or run by looking at the pace graph. If it was a bike activity the pace graph tends to be smooth. If it was a run the pace graph on Strava tends to be quite noisy. I am not sure this variance of speed alone could be used as a factor, but I think it could be a contributing factor.