Documentation for configuration or expalantion

deekerman commented

2026-02-28 01:17:11 -05:00

Owner

Originally created by @argasmail-cyber on GitHub (Dec 16, 2025).

Could you please let me know where I can find the documentation that explains all the configuration parameters?
I am particularly interested in the sections related to notifications and image capturing on trigger, as there are several options available. Unfortunately, the question mark icons do not provide any useful information.
I look forward to your feedback.
Thank you and best regards,

P.S. I am using visioneye in home assistant

Originally created by @argasmail-cyber on GitHub (Dec 16, 2025). Could you please let me know where I can find the documentation that explains all the configuration parameters? I am particularly interested in the sections related to notifications and image capturing on trigger, as there are several options available. Unfortunately, the question mark icons do not provide any useful information. I look forward to your feedback. Thank you and best regards, P.S. I am using visioneye in home assistant

deekerman added the

question

label

2026-02-28 01:17:11 -05:00

deekerman commented

2026-02-28 01:17:14 -05:00

Author

Owner

@MichaIng commented on GitHub (Dec 16, 2025):

Our wiki covers some of them, but a lot is missing: https://github.com/motioneye-project/motioneye/wiki/Configuration

What information are you missing exactly, how can the tooltip be enhanced?

motionEye in Home Assistant might work a bit differently, but the notification triggers should work the same way. I guess we can enhance the tooltips on the one hand, and create a new wiki page for all motionEye GUI settings with some more detailed info, where applicable.

@MichaIng commented on GitHub (Dec 16, 2025): Our wiki covers some of them, but a lot is missing: https://github.com/motioneye-project/motioneye/wiki/Configuration What information are you missing exactly, how can the tooltip be enhanced? motionEye in Home Assistant might work a bit differently, but the notification triggers should work the same way. I guess we can enhance the tooltips on the one hand, and create a new wiki page for all motionEye GUI settings with some more detailed info, where applicable.

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@argasmail-cyber commented on GitHub (Dec 16, 2025):

Thanks for the reply.
My needs are related the setting of the image to save, because if I set save picture with motion trigger the motion eye will sent a lot of photos. And i will receive via telegram a lot of photos. (Also I don’t understand the setting of the second of pictures in telegram configuration)
If i set motion trigger one photo I will receive via telegram only the message but no picture
I hope that someone can help
Regards

@argasmail-cyber commented on GitHub (Dec 16, 2025): Thanks for the reply. My needs are related the setting of the image to save, because if I set save picture with motion trigger the motion eye will sent a lot of photos. And i will receive via telegram a lot of photos. (Also I don’t understand the setting of the second of pictures in telegram configuration) If i set motion trigger one photo I will receive via telegram only the message but no picture I hope that someone can help Regards

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@MichaIng commented on GitHub (Dec 16, 2025):

In Telegram notification settings you can define a timespan, and it appends all pictures taken within this timespan around the timestamp of the motion detection. motion saves either a single frame or all frames of the motion event as pictures.

So if you want a single picture only, change the still images trigger setting to "motion triggered (one picture)". Else you can reduce the number of pictures by adjusting the timespan in the Telegram notification settings. I just see the tooltip says "interval" there, which is the wrong word. It is a timespan: pictures generated within a that timespan before and after the motion event timestamp are appended to the notification: https://github.com/motioneye-project/motioneye/blob/main/motioneye/sendtelegram.py#L73-L77

You could also reduce the frames per second processed by motion, but that can make motion detection less accurate, and you might want to be able to review more images if the excerpt sent to Telegram does not give enough hint.

I guess an option to append just a single image to notification messages, which generating more at the server, would not be a bad idea.

@MichaIng commented on GitHub (Dec 16, 2025): In Telegram notification settings you can define a timespan, and it appends all pictures taken within this timespan around the timestamp of the motion detection. `motion` saves either a single frame or all frames of the motion event as pictures. So if you want a single picture only, change the still images trigger setting to "motion triggered (one picture)". Else you can reduce the number of pictures by adjusting the timespan in the Telegram notification settings. I just see the tooltip says "interval" there, which is the wrong word. It is a timespan: pictures generated within a that timespan before and after the motion event timestamp are appended to the notification: https://github.com/motioneye-project/motioneye/blob/main/motioneye/sendtelegram.py#L73-L77 You could also reduce the frames per second processed by `motion`, but that can make motion detection less accurate, and you might want to be able to review more images if the excerpt sent to Telegram does not give enough hint. I guess an option to append just a single image to notification messages, which generating more at the server, would not be a bad idea.

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@argasmail-cyber commented on GitHub (Dec 16, 2025):

Thanks for your reply.
A useful improvement could be to send only one image via Telegram when motion is detected. On the server side, having a series of photos is very helpful, especially since we only store them for one week.
Another possible enhancement would be to analyze the photos sent via Telegram (those with clear motion) using LLM Vision in Home Assistant. Currently, I have linked MotionEye with LLM Vision through a webhook and stream analyzer, but on my hardware this process takes a few seconds, which means quick motions can be missed.
For now, I’ve set the Telegram field value to 3 seconds, and I’ll test it tomorrow morning. Let me know if you need feedback or additional tests.
Regards

@argasmail-cyber commented on GitHub (Dec 16, 2025): Thanks for your reply. A useful improvement could be to send only one image via Telegram when motion is detected. On the server side, having a series of photos is very helpful, especially since we only store them for one week. Another possible enhancement would be to analyze the photos sent via Telegram (those with clear motion) using LLM Vision in Home Assistant. Currently, I have linked MotionEye with LLM Vision through a webhook and stream analyzer, but on my hardware this process takes a few seconds, which means quick motions can be missed. For now, I’ve set the Telegram field value to 3 seconds, and I’ll test it tomorrow morning. Let me know if you need feedback or additional tests. Regards

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@MichaIng commented on GitHub (Dec 20, 2025):

LLM analysis is out of scope, a different topic amyway, and for motion, not so much for motionEye IMO, basically as an altwrnative to the static "changed pixels across X frames" method currently used.

But I agree that "send one picture" would be a good addition for notofications. Instead of a timespan, some offset could be defined, or we just use a similar method used to choose the single frame to store at the server side, or the offset used to generate a preview image from a movie file.

@MichaIng commented on GitHub (Dec 20, 2025): LLM analysis is out of scope, a different topic amyway, and for `motion`, not so much for motionEye IMO, basically as an altwrnative to the static "changed pixels across X frames" method currently used. But I agree that "send one picture" would be a good addition for notofications. Instead of a timespan, some offset could be defined, or we just use a similar method used to choose the single frame to store at the server side, or the offset used to generate a preview image from a movie file.

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@argasmail-cyber commented on GitHub (Dec 20, 2025):

Thanks for your reply so I will wait the update for a single foto to the message.
An other improvement could be to use this foto to send to to other service or to the calendar as event.
The possibility to use the single photos for ia analyzing could be useful after the use of webhook but I don't know how to search automatically the last picture name as variable. Like take the last single photo taken and send to other services.
Regards

@argasmail-cyber commented on GitHub (Dec 20, 2025): Thanks for your reply so I will wait the update for a single foto to the message. An other improvement could be to use this foto to send to to other service or to the calendar as event. The possibility to use the single photos for ia analyzing could be useful after the use of webhook but I don't know how to search automatically the last picture name as variable. Like take the last single photo taken and send to other services. Regards

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@MichaIng commented on GitHub (Dec 20, 2025):

Yeah one by one. Once we have a logic to catch one frame of a set of local frames/pictures from a motion event, we can think about making use of that logic in other actions, e.g. make it available with some special specifier for the webhook URL or so. Though currently not sure how to upload an image without larger rework, not sure whether this can work/is commonly possible as POST payload? Anyway, I personally have limited time and rather focus on fixing things and enhancing/hardening existing methods, than implementing new features.

@MichaIng commented on GitHub (Dec 20, 2025): Yeah one by one. Once we have a logic to catch one frame of a set of local frames/pictures from a motion event, we can think about making use of that logic in other actions, e.g. make it available with some special specifier for the webhook URL or so. Though currently not sure how to upload an image without larger rework, not sure whether this can work/is commonly possible as POST payload? Anyway, I personally have limited time and rather focus on fixing things and enhancing/hardening existing methods, than implementing new features.

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@argasmail-cyber commented on GitHub (Dec 21, 2025):

Thanks for any help in case of help tell me to test and see features I can do also simple tasks
Thanks again regards

@argasmail-cyber commented on GitHub (Dec 21, 2025): Thanks for any help in case of help tell me to test and see features I can do also simple tasks Thanks again regards

deekerman commented

2026-02-28 01:17:15 -05:00

Author

Owner

@ToeKneeFan commented on GitHub (Feb 21, 2026):

@argasmail-cyber You mentioned wanting to keep multiple images in a series server side but only sending one via Telegram. Couldn't you just save only 1 image ("Motion Triggered (One Picture)") and then have your "multiple pictures" be the mp4 video?

That is essentially what I do. I save 1 picture on motion. It also continues to record video.

I then have a File Storage > Run a Command script that checks whether a created file is a jpg. If it is, it sends the image via email as an attachment. (In your case, you could replace with a script using a Telegram CLI.)

@ToeKneeFan commented on GitHub (Feb 21, 2026): @argasmail-cyber You mentioned wanting to keep multiple images in a series server side but only sending one via Telegram. Couldn't you just save only 1 image ("Motion Triggered (One Picture)") and then have your "multiple pictures" be the mp4 video? That is essentially what I do. I save 1 picture on motion. It also continues to record video. I then have a File Storage > Run a Command script that checks whether a created file is a jpg. If it is, it sends the image via email as an attachment. (In your case, you could replace with a script using a Telegram CLI.)

Rows
Columns

Documentation for configuration or expalantion #2669