A claimed confrontation outside of a courthouse leads to a restraining order against a man. There's surveillance video that the man believes will exonerate him, but the court rejects the video in its entirety.
Here's the case out of California: A man is accused of stalking a woman and making violent threats. They go to a trial for a restraining order and after the order is granted, she and her friends say the man yelled and threatened them from his car outside of the courthouse. The woman, a friend of the original plaintiff, ran back into the courthouse out of fear and later got her own (second) restraining order against him.
*** *** ******** ****** *** ******** never ******** *** ***** *** * new ***** ***** ** *** ******** -- ************ ******* *** **** ***** tower ****. ** **** *** ******* would **** *** ******** ***** ********, and **** ***** ***** **** ***** show **** ** *** *** **** the ********** *** **** ** *** on *** ***** ** *** **** the ******** ***** **** ********.
***** *** ** ***** **** ******* on *** ******** ** *** **********. Here *** *** ** ****:
******* ****, ***** ******* ******* **** courthouse *******, ** ********* ***** ******** to *** *** ********* ******* ** the **** *** *** *** *********** orders ***** ******:
"*** ***** *** ****** ***** ****** and, ***** *******, ***** ****** ********* show * ****** ******* **** *** street *** *** **** **** ****** they ***********. *'* *** **** **** the ****** *** ***** ** ****** the ***** ** ****** ** ******* this *******. ** *** ****** **** it *** **** ****** ** ********* and ******** ******, *** *** ******, there ** ** ***** ** ***** tapes. ***** ***** *** ********** ** --it ******* ** ** **** ** what ** ********* ******* *** ********, but, *****, ** * ****, **** of *** ******** **** *** *** one ****** ** * ********, *** very **** ****** **** *** *** there. *** ** *'* *** **** how ******** **** *** *** ** the ****** ***** ** ********** **** can ** ****, **** **** ***** be ********** *** ******** ** *********** value. * ******* *** ***** *** look ** **, *** *** ***** probably ***** *** ** *** ***** what *** ******* ** *******. *** then **'* ** ************** *******."
***** ********? ******** ******* ** "****** disappearing" *** **** ********* ** *** ******* from *** ******** ******* ****. ** *** ********** **** ***** cameras **** *** ** ****** ** motion *** *** ********** ********, ***** is **** *** ** ************** [**** no ****** *********].
*******, *** ******* **** *** *** set ** ******-***** *********, ********* ** Lt. ***** **** ** *** *** Diego *******'* ******. *** ******* *** set ** ****** ************ *** ** the **** ** *** ********* *** haven't *** *** ********* ****** ********. Although, ** ****, *** ********** ** bound ** **** ******** **** **** to ****.
*************, *** ***** ******* ** *** available.
We usually recommend clients test viewing recorded video and exporting video from time to time. You can try and come up with all the monitoring systems you want, but actually use is the best assurance the system is working properly.
Would be nice to know if the courthouse security had a procedure for intermittently testing the system or a service contract that included it.
In this case, it does not seem the judge is contending there was a technical issue, as the Sheriff's Office confirms.
Since the defendant was arguing that no confrontation occurred, you'd have to have comprehensive coverage of everywhere to show that. Otherwise, one could still contend that, yes, the confrontation occurred but the cameras missed it because they did not cover this area. Yes/no?
To play devil's advocate against myself, presumably the plaintiff would need to testify where and approximately when the confrontation occurred. To that end, if the video was available at that time and location and did not show it, that should be sufficient, no?
Doesn't this head down a similar path to radar equipment...when was it last calibrated...by who..are they certified....etc....any attorney could pick it apart
Proving something didn't happen is much harder than proving something did.
He would need a continuous video during the entire time of the alleged incident. The judge's description of the video doesn't give the impression that it is continuous (although it may very well be).
The burden should be on the plaintiff to prove it happened, but not so much when it comes to a restraining order.
A restraining order, in and of itself, would be no big deal. Just stay away from her! The problem is having one against you can affect you in so many ways. For instance, federal law will prevent you from owning a gun while you are under a domestic restraining order.
Snowman
Judge:
"I suppose the court can look at it, and you could probably point out to the court what you believe it depicts. And then it's an interpretation process."
ummm.... isn't that what courts do? Interpret evidence? Why would 'having to interpret' actions on a submitted video be a basis for exclusion? And what about the cell-tower tracking evidence? Judge didn't say anything about it...
The judges reasoning is technically flawed, but because the evidence itself is being excluded, the defendant can not even challenge the original decision.
The judge was probably just annoyed that some idiot appealed a restraining order so he refused to allow it - even if his reasoning was flawed. Trivial crap like this aint headed to the Supreme Court.
Note: I agree with Snowman - just leave her alone, man! :)
It appears as if the primary reason for not admitting the videos is that they do not contain clear evidence. The fact that the videos may have been the result of motion-based recording, or some other setting or problem that caused them to be non-continuous is really of secondary concern.
People will point to cases like this and say that motion-based recording is inadmissble, or draw some other flawed conclusion. But the main issue seems to be that there is no evidentiary value in the videos.
If a passer-by happened to take a single, clear photograph of this man harrassing someone, that image would likely be admissable as evidence. Proving his action, or location or both. Image evidence (which is what video is) is not a factor of compression schemes, continuity or anything else. It's a factor of does the data you have provide clear benefit to the prosecution or defense?
Would a continuous recording have been admissable? Possibly, but not because it was continuous, but more because it would have left less open to interpretation. Does enabling continuous recording on your NVRs increase the chances that the video is admissable? Most likely the answer is yes, but that is because you get more data (evidence) and have less gaps that can be twisted into arguments that the data is not conclusive.
IANAL.
The appellate judge is not allowing the video evidence in. So there is not going to be any arguing about whether or not the data is conclusive in this case.
The judge appears to be using the 'disappearance' of cars and whatnot on the video to nullify the technical merits/validity of the video itself - allowing him to block the introduction of this 'evidence' outright.
Who knows what the video shows if the defendant is not allowed to use it in his argument? This is my only issue with the whole thing.
This points out the challenge of proving a negative. In many cases, it simply can't be done. This is one very important reason that our founding fathers instituted the presumption of innocence.
IANAL
Absense of evidence is not evidence of absense. If the video had been continous (contiguous) with no gaps, and covered the area, it might prove that the accused was not present. As there ARE gaps, it can't, and thus is not suitable to use as "evidence of absense".
If, instead, the accused could acquire video from a system at his alledged location, and that video showed him there, irregardless of gaps, it might be admissable.
I am not sure that in a case regarding a restraining order (the 2nd), that it would be severe enough to get a court order for the phone company to surrender tower information. And even if they did, it would only prove the location of the phone. Not of its owner.
I just now saw the Kendrick Johnson case, which is a lot more tragic. It's mind-boggling to me that you would spend a small fortune on a surveillance system that only "just might work" when you need it. The CNN report alludes to the idea that someone tampered with the evidence. That's obviously a lot more dramatic and exciting than saying that the system was shit not too "user friendly".
As it turns out, running the systems with no time sync, 1 fps, and poor motion detection setup made it real hard to determine what went on when something did go down. The lack of time sync and 1 fps are smaller problems than the poor motion detection setup, which leads me to one of my pet peeves :
There are too many damn options!
Each option is a small trap for the integrator to fall into. And to make matters worse it's really difficult to validate the settings. Anecdotally, I spoke to an integrator recently; he always sets up for continuous recording, because he doesn't have time to sit there an tweak the motion detection settings properly. I think he should be going from ultrasensitive to insensitive aiming for just above "too sensitive", but what if you don't, what if you go the opposite direction and stop just as it triggers on some random person walking by at that instant. How do you know if someone walks by and don't trigger motion. How do you know if it picked up every single person that walked by 24/7, 52 weeks of the year. Taking rain, snow and fog into consideration. What about people moving slowly, or fast and so on.
Then when you are done playing with motion detection, you can play with the quality settings, bandwidth settings and so on, and hopefully you don't cranck up the compression to squeeze in more days on the harddrive, while making the image quality so piss poor bad that you might as well have installed a 640x480 USB camera from 2002.
With that said, Happy New Years folks!
Great points! On the motion detection issue, generally multi-streaming refers to the ability to render streams in different resolutions or formats. However, during a system's commissioning, there is an argument for a different multi-streaming approach: streaming continuous video and streaming only motion-detected video. While there's no magic bullet, given the state of the art today, is there any merit to such an option to verify performance in the first week or so of operation (or any time doubt is raised), before "settling down" to only the storage-efficient long-retention motion detection mode?
Another potentially interesting approach might be as follows:
We talk about server side or camera side motion detection as if they're mutually exclusive. Suppose you dual stream continuous and on-motion video to the VMS. Then the VMS motion-detects on the continuous video. Using the video time stamp as determinant, the server can flag and store on either of two conditions: server detects motion the camera missed, or camera detects motion that the server missed. Given that the two likely use different algorithms, they won't agree perfectly so the filter should have some tolerance for mild temporal error, but significant on-going mis-match should be a flag that one or the other is not set up properly. After a week or so of running like this, again one would likely wish to fall back to the lower resource configuration of single vice dual motion detection.
Quite aside from the wide range of performance that IPVM exposed in recent motion detection testing, both of these suggested approaches also suffer because redundancy increases resource requirements (and cost). One problem is, how does the user attain confidence in finicky and difficult to set-up functions? Should the installer loan a temporary adjunct processor for secondary motion detection, only during the commisioning period? If the motion detection capability is important, it seems unfortunate that our tools for assessment appear to be relatively weak.
I guess this rambling note had brought me around to your integrator's mind-set: "he always sets up for continuous recording... ." Why add all this complexity and additional resource usage? What is motion detection about? I can think of three uses: extending storage intervals in a given size datastore, limiting bandwith usage on austere (eg wireless) links, and making video more easily searchable or accessible. The latter two seem marginal to this discussion, so I'll focus on affordable resourcing. If commissioning or implementation approaches for motion detection burden the system or installer with additional costly resources, it's probably better to have just purchased additional storage and record continuously and have greater confidence that, when needed, the video will be there.
Thanks again, IPVM, for highlighting a really significant issue that we don't necessarily appreciate when we think we're making wise procurement, design, and set-up decisions.
