Rain and Flying Insects

[pdownes2001_384636162646205]

I'm plagued with false alerts from flying insects and rain.
Would it not be possible to tweak the software/firmware so that users could define a Minimum Length of Motion?

A flying insect presents an event that lasts less than a second. Raindrops even less. If users had the option to set their cameras so that they ignore such short-lived motion events, false alerts could be much reduced.

[mick94115_491629178327178]

Hear hear!

This feature is absolutely needed. At night, when I do want alerts, I need to turn down the sensitivity so low, I'm sure it would not detect a person in my back yard. Even if I can have a feature to "Ignore X number of sequential events before triggering alarm", this way when a bug flys by, it will ignore it as opposed to a person, who is probably going to be on camera longer.

[Reolink Fiona]

@mick94115_491629178327178
Thank you for your feedback. I have collected your problems and we will seek more ways to figure out the problem. Now you can try to use the object size and the alarm delay to filter some false alerts now. Lower the motion sensitivity and try to use these features to filter the motions.

[Crimp On_62210811129]

I have mentioned this several times. It makes SO much sense.

I also think there is no reason not to have SETTINGS for several parameters.
"Prerecord", for example, seems to be a fixed number of seconds or none at all.
The user should be able to set up to some maximum value, say 0-20 seconds.
Post record is the same.

I finally solved the "flying insect problem" by purchasing separate Infrared LED lights and placing them 6-8 ft. away from the camera. Bugs are drawn to them and do not appear to be BIG BRIGHT OBJECTS to the camera.

[Josh_576350005272911]

@crimp-on_62210811129 This seems like a good idea! Might have to try this.

[pdownes2001_384636162646205]

I'm glad I'm not alone on this. The now obsolete Ucam247 and Phylink cameras had those options and they were probably the most useful options for avoiding false alerts and ensuring a real event was properly saved.

Did anyone from Reolink add a comment?

I can't see it being a big problem to tweak the software and the benefits would be huge.

[Cynthia_124785627824270]

We really appreciate your feedback.
Please allow me to explain briefly how the camera detection works: the camera detects the motion based on the image changes (the background) in each frame. So when the object reaches a certain degree the camera will be triggered.
In order to reduce false alarms, it's advised to combine the methods of setting up the proper sensitivity and defining proper motion detection areas on Reolink Client software.
Intensive motion like rain/snowflakes, insects attracted by IR lights, may also cause frequent alarms which may not have a good solution to it.

[pdownes2001_384636162646205]

We really appreciate your feedback.
Please allow me to explain briefly how the camera detection works: the camera detects the motion based on the image changes (the background) in each frame. So when the object reaches a certain degree the camera will be triggered.
In order to reduce false alarms, it's advised to combine the methods of setting up the proper sensitivity and defining proper motion detection areas on Reolink Client software.
Intensive motion like rain/snowflakes, insects attracted by IR lights, may also cause frequent alarms which may not have a good solution to it.

Cynthia
Thank you for your input.
I have many years experience as a user of security cameras. I was also an engineer for 42 years so I understand the reluctance to accept that someone else might have a solution for a problem.

I would be very grateful if you at least forwarded the proposal to introduce a 'Length of Motion' feature.

In my experience, it was the single most effective solution for reducing false alerts caused by rain and insects.

It actually works very well and it is not hard to implement.




image_2021-05-13_102133.png

[Jose_320595376935068]

In order to reduce false alarms, it's advised to combine the methods of setting up the proper sensitivity and defining proper motion detection areas on Reolink Client software.
[/quote]

Although tweaking sensitivity may do the trick in some situations it is not by no means a solution because it also affects distance and on time catching of real intrusions, so defeating the purpose of having a camera to detect them as soon as possible. The same is true for detection areas, of course very useful to mark where you don't want to get an alarm, but no at all to discern about what is causing it.

[quote quote=1458619]
Intensive motion like rain/snowflakes, insects attracted by IR lights, may also cause frequent alarms which may not have a good solution to it.

In this and previous threads you have already received an excellent idea to largely reduce this problem.

[Crimp On_62210811129]

This is indeed the "crux of the matter". The current program compares one frame to the NEXT frame and calculates how many pixels have changed. If the number of changed pixels is greater than the cutoff value, by definition there is "motion".
The proposal is to make the calculation more complicated. That is, to COUNT the number of frames that are different rather than trigger motion on one change. For example, every time one frame is different than the previous frame, increment the counter by one. If the new frame is not different than the previous frame, reset the counter to zero. When the counter reaches some value, declare "motion". Suppose a user says, "I do not want to see 'motion' that lasts for less than 2 seconds." Then the number of successive frames that change has to reach 2 times the number of frames per second. (At 20 frames per second, that would be 2*20 = 40). If there are 40 frames in a row that are 'different', then something I care about is 'moving'.

I understand this is NOT EASY. An insect may not "move enough" from one frame to another to meet the criterion.

But, that is the whole point of CHALLENGING programmers. They are SMART. They will find a way.

[Cynthia_124785627824270]

Thank you for your suggestion, we will forward it to our R&D team. Hope that they can make better products in the future!

[pdownes2001_384636162646205]

Thank you Cynthia. Please keep me posted when they have had a look at it. I would like to see some progress on this.

[Crimp On_62210811129]

I volunteer to be a Beta tester if this capability can be developed. I understand testing will be a chore because camera firmware has to be changed back and forth. I'd suggest developing for the camera web interface first and add functionality to the Mac/Windows Client after the firmware works on the camera.
I have RLC-410-5MP, RLC-420-5MP, RLC-510WA. Would even buy a 4K camera if the developers want to start at the high end.

On that screen shot of the PHYLINK camera, I am confused by the options:
Minimum length of motion. To me this means "if motion does not persist longer than this, I don't want to know about it."
Minimum ignore time of motion. This sounds like saying the same thing, only backwards.

[pdownes2001_384636162646205]

I remember being baffled by that second option too.

I wasn't the only one and several people did ask the question.

Sadly, the explanation made no sense. The Phylink guy who provided it did not have a very good grasp of English.

I never used the option.

It's still a mystery.






no-one-knows.gif
no-one-knows.gif

[Cynthia_124785627824270]

Thanks for your feedback! Your request has been collected. Engineering is reviewing the feasibility of implementing this idea. Thank you sincerely for sharing your thoughts. We will take your request seriously and work hard to make our products better~

[pdownes2001_384636162646205]

Thank you Cynthia.
I hope they find a way to make it happen. It will be a game-changer for us and a great marketing ploy for yourselves.

False alerts are probably the most complained about things for all brands of security cameras. You could steal a march on the rest with this one.

[pdownes2001_384636162646205]

Hi Cynthia.

Have you any news from Engineering about Minimum Length of Motion?

[Cynthia_124785627824270]

Hello! We have forwarded the feedback to our engineers, it will take time to evaluate the request and to implement it.

[pdownes2001_384636162646205]

Hi again Cynthia.

I have just noticed that a new feature in the latest iPhone client app is described as,
"Add tolerance settings to set up how long a detection object is allowed to stay in detection area."



This sounds very much like the Minimum Length of Motion feature that we have been talking about.

However, the new feature is only available for some devices.

Please can it be made available for other devices such as the RLC-520 and RLC-410-5MP via a firmware update?





image_2021-06-14_014039.png

[Cynthia_124785627824270]

Hi, thank you for your suggestion! We will forward your suggestion to our product team.

[pdownes2001_384636162646205]

Thank you Cynthia.

Could you tell me which devices are compatible with the new feature please?
I might be tempted to buy one or two.

[Cynthia_124785627824270]

Hi, this feature may not be applied to the camera now. Once there is any update, we will let you know. Thank you for your interest!

[pdownes2001_384636162646205]

Is it likely to happen withing months or years?
I am close to making a decision on my next camera purchases.

[Cynthia_124785627824270]

Hi there, this feature will be available with the cameras that the camera name is with A. For example, RLC-811A, RLC-1210A,etc.

[trelower2_471448109822110]

@cynthia_124785627824270 Any news on this upgrade?

I am a new user of RLK8-820D4-A and am already sick to death of the false alarms from rain. So much so that I have had to turn of "push notifications" making the system as much use as an ashtray on a motorbike!!

If this is not going to be sorted then I will return the whole lot as unfit for purpose which would be a shame as apart from this the system is quite good

[Reolink Lorenz]

@trelower2_471448109822110 Hello there, we are still testing the latest firmware for the smart detection lineup. If you want to try the firmware now, please submit a request here, https://support.reolink.com/hc/en-us/requests/new, and ask our team to send you the latest firmware.

[pdownes2001_384636162646205]

Thank you Cynthia.
That is great news. I have plans to deploy cameras in my garden to watch out for rats and other vermin.
The length of motion feature is critically important.
I look forward to the feature being rolled out. I am sure it will be a success.