ZoneMinder Forums

Hi all,

I picked up a couple of relatively cheap b&w outdoor cams to monitor down the side of my house. Unfortunately I installed the cams (dragged cabling through the root etc) and setup the cams to find that they seem to be rather *noisy*

I have to qualify the word noisy. The picture looks very good if a little low res (352x280 is the most I can get out of them). However whenever I configure an active zone I start receiving alerts almost continuously - I look and can not see the movement, but it is alerting.

I have posted one such frame with the 'supposed' movement outlined by the analyser:



There is no notable change in the frame. I have set very high values for sensitivity and still find at times the entire zone is alive with supposed movement, with nothing notable visible in the frame.

My final test was to connect a second camera alongside this one. The second camera is working as expected and is only triggering when there is movement as it should.

Has anybody come across anything wierd like this before ?
Would noise the a likely candidate even if it is not visible ?
Any suggestions or should I continue with my plan to return the camera's ?

Cheers
Jason.

What's the sensitivity on the zones?

I'd guess you've got a video signal level that's fluctuating. This could be caused by all sorts of things, like:

* signal loss in a long video co-ax run
* interference from (say) a mains cable and a poorly screened video cable or connector
* the power supply to the camera not being steady
* etc.

You could try turning the pixel threshold sensitivity of the zones down a bit, perhaps.

You could also try changing the rate of reference image blending.

Other ideas, anyone?

I have tried a variety of settings and while the occurrences of false alarms is reduced, there are still instances when movement is detected with the blob size over 90%. Essentially - not overly useful for motion capture.

To test I wired up a second camera I had (little colour jobby) and it is working perfectly with alarms being triggered with any movement around my tool shed and the back door to the house.

To further test the camera I installed it in my computer room removing the chance of line noise. Events were still being detected frequently even with all lights off and the door closed (no movement). Reducing the brightness a *lot* also improved the alerting, but not to the point it was useful. Increasing the blend percentage also helped.

Essentially - I bought cheap camera's and got what I paid for...

These cams are no good for motion capture as the signal noise seems to be inherent in the device, I will be buying some better cams for monitoring outside and will probably use these two camera's for full time recording of my front and back yard where there is far too much motion anyway (light reflecting off the pool, shadows etc).

Cheers
Jason.

It looks like playing with the contrast and brightness may have resolved the worst of the false positive problems with this camera...Under test conditions anyway. I'm going to reinstall this camera outside and see if it is more reliable with the new settings.

Cheers
Jason.

I am having the exact same problem. Please let me know exactly what the changes are that you made and whether the changes were succesful or not.

Maybe we should start a thread as to what are good cameras and what are not? I know that I have purchased 6 cameras so far with only two usable since the variety of the other cameras are either not as low-light sensitive as advertised, they wash out in bright light and all have extremly noisy images.

Thanks
Skydiver