is i9-9900K still the king?

bp2008

Staff member
Joined
Mar 10, 2014
Messages
12,666
Reaction score
14,006
Location
USA
Also somewhat related, from the Blue Iris help file:

Video analysis

By default, Blue Iris analyzes 1-2 frames per second of video from each camera to look for
motion. The algorithms employed are fairly simple and not in general CPU-intensive.
However, when motion is detected, the rate of analysis goes up to at least 8 frames per
second in order to more accurately follow objects. If you do have a number of cameras with
a considerable amount of activity, this may begin to contribute (still in a small way) to the
overall CPU demand of the software. There are a couple of ways to mitigate this:

• Do not use the “high-def” setting on the Motion detector page. This uses up to 4x the
number of pixels as input to the algorithms.

• Don’t use the “Gaussian” algorithm.

• Consider using camera-based motion detection if offered by your camera. This does
require that the camera supports and is configured to use ONVIF GetEvents via
PullPointSubscription and that the software understand the type of events the camera is
sending (this is not standardized). You can view what the camera sends in response to
motion or other triggers using the Events page in the ODM (ONVIF device manager)
software.
 

abo

Young grasshopper
Joined
Mar 21, 2018
Messages
44
Reaction score
2
IMHO motion detection is so 2019... AI Tool or AI in general, is the future...

View attachment 75057
Yes, I totally agree but do you have anything usable for normal home users that is not some kind of cloud service?
I wish the software of the self driving Teslas woud be open source.
That code would be a good starting point for this usecase also.
 
Top