I did the 3x factory reset, and am trying a couple of other things. I was using h.265 before, now running h.264. I had the picture flipped, leaving it upside down for now. If it doesn't work any better now, I'll kick it up to Andy. I just need to herd some cats through the tripwires, and there's more snow starting, so it may take a few days.
[Partially Solved] Lame IVS on IPC-T5442TM-AS
- Thread starter tigerwillow1
- Start date
After 3 factory reset should I set up the configuration from scratch or is an import from the saved configuration successfully?
I will try it next few days.
Intrusion and loitering works very well, but tripwires very bad.
I discovered another issue with my 5442TM: The IVS settings don't work via the 5216-4ks2 NVR. They do work with all my other cameras. Andy says that there's no newer firmware available past the .10 version. I also tried older firmware versions with no improvement. So for me, IVS is unusable and I can only hope that some future firmware will fix it. The irony is that I'll likely keep the chinese market camera the 5542 is replacing set up beside it. The 5442 will give me the better image, and the chinese camera will give the IVS. The NVR is able to map one camera's IVS trips onto another camera's timeline. Another solution would be to give in and switch to BI. I can't add any more 5442's until something changes.
In various threads some people have reported IVS problems with the 5542 while others say it works great. My 5442 had the yellow triangle license issue. Could there be any association between that and IVS problems even after it's fixed? That could be answered with data from 5442 owners posting how well their IVS works and if they did or didn't have the yellow triangle problem.
For me, it's: Had the yellow triangle, IVS doesn't work very well.
In various threads some people have reported IVS problems with the 5542 while others say it works great. My 5442 had the yellow triangle license issue. Could there be any association between that and IVS problems even after it's fixed? That could be answered with data from 5442 owners posting how well their IVS works and if they did or didn't have the yellow triangle problem.
For me, it's: Had the yellow triangle, IVS doesn't work very well.
Exactly the info I was looking for. I just can't find any settings, resets, reboots, or magic incantations to get the IVS to detect anything smaller than a person or deer. I think I need to quit throwing good hours of experimenting after bad and just give up. Here's an example picture of a cat walking right through a tripwire undetected. No difference if it's day or night. Even my junk chinese cameras will trigger on a mouse in that location. The object in the upper right is the chinese market camera I thought I was replacing.
You may try one zig zag line/rule instead of multiples...
Are you setting IVS and Smart Plan on the camera or the NVR? I found setting it on the camera works reliably.
Are you setting IVS and Smart Plan on the camera or the NVR? I found setting it on the camera works reliably.
Agree with @bigredfish... for an area like this I use an Intrusion zone in the camera IVS. Mark it for entry/exit and set the max/min pixels. Works great on my 5231s.
You definately dont have the options for "human" and/or "vehicle" checked with these IVS rules? If you do, then it will of course, ignore that cat. I'm presuming the AI in these cameras is basic size/shape detection (hence the need to calibrate).
I use intrusion zones too. They have an option for “appears”, which covers the situation where the object made it across the edges before the camera recognized it.
Thanks for all of the good feedback. I've tried everything mentioned in the last few posts except for zigzag tripwires. I don't want to fall into the trap of feeling infallible, but I've got probably 20 hours into trying various combinations of settings. With any of my other cameras, I can have IVS set up and working right in 10 minutes. What I see is very consistent with the human object filter being turned on. The camera will draw green boxes around people and some deer, nothing smaller than that. For all the time I spent I could have set up a BI system. I've been resisting. The Dahua NVR is reliable as heck, but it too is full of buggy firmware. Here's a fox that slipped through early this morning, detected by IVS on a different camera.
sebastiantombs
Known around here
I've had mixed results with tripwires and had eliminated them. After reading this mornings posts I added a zone and set it for intrusion with both "appears" and "cross" checked. Works like a champ, so far with some testing.
sebastiantombs
Known around here
I pulled the 5442 into the house and gathered more data. No IVS improvement. Went back to the totally reliable hacked chinese camera outside. I spent all afternoon trying every combination of settings I could think of, 3 different firmware versions, reloading the yellow triangle license, and IVS works exactly the same in every case. So far, I've been denying that it could possibly be a hardware issue because I figure all of the IVS is firmware driven, but maybe there's a video processing chip that could be bad or something like that. I managed to make a list of (I assume) unrelated bugs while experimenting. First a picture. There's an intrusion box and 2 tripwires set. I could not cause an IVS trip moving the envelope anywhere in the frame. The only way I could cause an IVS event was to stand up in front of the camera. The bug list follows the picture.
The bug list:
1. IVS events don't go into the camera log (by comparison motion detect events do)
2. When camera switches from night to day, a video tamper event is logged
3. IVS works with smart codec on. Is this a new feature???
4. When using my Firefox, I believe without a plugin, the images in the IVS setup screen are delayed by 10 to 60 seconds. When using Pale Moon with plugin, there's no delay.
The bug list:
1. IVS events don't go into the camera log (by comparison motion detect events do)
2. When camera switches from night to day, a video tamper event is logged
3. IVS works with smart codec on. Is this a new feature???
4. When using my Firefox, I believe without a plugin, the images in the IVS setup screen are delayed by 10 to 60 seconds. When using Pale Moon with plugin, there's no delay.
Current status is I'm going to check with Andy about replacing the camera, which isn't happening until the other side of Chinese New Year.
I added detailed info to the first post of the thread, am making this post to bump it to the top. Bottom line: The only firmware version that gives usable to me IVS is V2.800.0000000.8.R , and I won't be able to get a replacement camera until the coronavirus disruption has passed.