MagicMirror Forum

majik655

@ghiesseldorf said in MMM-Facial-Recognition-OCV3 FATAL: OpenCV Major Version must be 3:

MMM-Facial-Recognition-OCV3

This won’t help you with this module but this module supposedly is old and has issues so this says…
https://forum.magicmirror.builders/topic/10674/mmm-face-reco-dnn-all-new-face-recognition

This module however works fine (I just installed last week)…
https://github.com/nischi/MMM-Face-Reco-DNN

majik655

Not sure if this will help you and sorry if this is obvious…
but if you want you could pre set up their wifi for them…

edit the wpa_supplicant.conf

When you connect to wifi the wpa_supplicant.conf file is created with your ssid’s and passwords etc etc…

So you can connect to your own wifi… then go edit the conf file and ADD their wifi …

When you first write the image to the sd card…there is no conf file because you havent connected to wifi yet (obviously)… so when done writing your sd card image… you can open the sd card (on widows or whatever) and create the wpa_supplicant.conf in the root with the info needed … otherwise just connect to wifi then edit the file.

Not sure if that helps you all that much but you can pre set it up you just need their info like SSID they would connect to and password

sudo nano /etc/wpa_supplicant/wpa_supplicant.conf

Below is what mine looks like with 2 ssid’s to choose from …

ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
update_config=1
country=US

network={
        ssid="SSID HERE"
        psk="PASSWORDHERE"
        key_mgmt=WPA-PSK
}

network={
        ssid="SSID2 HERE"
        psk="PASSWORDHERE"
        key_mgmt=WPA-PSK
}

majik655

@cmer3 said in Best PIR Strategy for RPI4?:

Hi, I am a noob , doing my first mirror project. I plan to have a PIR sensor to wake the mirror up from sleep. There seem to be 3 ways to do this: MMM-PIR Sensor, MMM-newPIR, and what I guess is a shell script? at https://www.putorius.net/using-pir-sensor-for-motion-detection.html. Is there a best approach for RPI4?

If it matters any… I used to use MMM-PIR but for the monitor I am currently using …when the monitor “goes to sleep” because of no motion, it would put my monitor to “sleep” …displaying a banner… not waking up… or you will get it to wake up but it shows the input HDMI that it is on (then goes away)…

With MMM-NEWPIR … it has the ability to put up a “black screen” type of thing… keeping the monitor “awake” but it is black… when motion it goes away and restarts the timer as motion continues…

That is basically the only reason why I used it… and I have it working on a PI3b+ and a 4 (4g ram).

majik655

Installed module last night … Great work! and Thank you again for creating this module!

While installing I now understand why the CODE from ecobee module may have been missed by others.
I was expecting to see the code and still took a minute or second look to see the code at the very bottom of the last sentence in the instructions on screen. If this CODE could maybe be larger or maybe a color that stands out?? Might help a little .

I too would like to choose which sensors are shown…or show them in a specific order.

One more thing that I would love would be making the module smaller.
I am still learning all of this and I played with Custom.css but of course the module will start to cut the wording and sensors off from visually being seen from the bottom up.

I will play more with it tonight… but this is awesome already! Thank you!

majik655

@cmer3 said in Best PIR Strategy for RPI4?:

Hi, I am a noob , doing my first mirror project. I plan to have a PIR sensor to wake the mirror up from sleep. There seem to be 3 ways to do this: MMM-PIR Sensor, MMM-newPIR, and what I guess is a shell script? at https://www.putorius.net/using-pir-sensor-for-motion-detection.html. Is there a best approach for RPI4?

If it matters any… I used to use MMM-PIR but for the monitor I am currently using …when the monitor “goes to sleep” because of no motion, it would put my monitor to “sleep” …displaying a banner… not waking up… or you will get it to wake up but it shows the input HDMI that it is on (then goes away)…

With MMM-NEWPIR … it has the ability to put up a “black screen” type of thing… keeping the monitor “awake” but it is black… when motion it goes away and restarts the timer as motion continues…

That is basically the only reason why I used it… and I have it working on a PI3b+ and a 4 (4g ram).

majik655

@ghiesseldorf
So I am not a developer nor programmer just a normal person who follows directions

So what I know is I followed this for the open cv stuff because I had issues following other tutorials or got lost from all the updates that have happened and the links saying to use a specific version yet the links provided point to a newer version…

In any case maybe this will help you…
In the modules installation instructions it gets to the point where you install Open CV…
And provides links to it depending on what operating system to install it…

SKIP over that
to where it says : OpenCV 4.1.2 quick installation (wthout compliation) for Raspbian Buster
And points to here: https://github.com/cyysky/OpenCV-Raspberry-Pi-4-Package-for-Python

I followed those steps, then continued on in the module’s install instructions…

So I am unsure where you are coming across that error (it sounds familiar, but I have done so much I can’t remember)… Installing openCV following those instructions is what made it “work” for me…

Also for the dataset it confused me at first and really all that is really needed is simple pictures from your phone (selfies) and put them in the correct directory as the instructions specify. (Anyway this was just a mention of what got me after the installation part of opencv but all works now I am also using a logitech usb c930 camera if it matters )

majik655

@ghiesseldorf said in MMM-Facial-Recognition-OCV3 FATAL: OpenCV Major Version must be 3:

MMM-Facial-Recognition-OCV3

This won’t help you with this module but this module supposedly is old and has issues so this says…
https://forum.magicmirror.builders/topic/10674/mmm-face-reco-dnn-all-new-face-recognition

This module however works fine (I just installed last week)…
https://github.com/nischi/MMM-Face-Reco-DNN

majik655

@Bugsounet said in MMM-GoogleAssistant with MMM-MotionDetector:

@majik655 said in MMM-GoogleAssistant with MMM-MotionDetector:

MMM-NEW-PIR

I have to update this module butI really don’t use it, because all fonctionnality is on A2D
I start to code A2D pir module to MMM-NewPir but not finish
I will continue it and commit it to master soon

Note:
For GA/A2D users it’s more natural to use screen/pir sub-module of A2D because it’s lock screen on GA/A2D use and can’t turn off
MMM-NewPir have not this fonctionality (and can’t never have it, because i don’t want to do a dual code)

example:

with GA / youtube and MMM-NewPir
defined tiner of NewPir is for example : 2 min
duration of video +/- 4 min
NewPir will turn off video at 2min

with GA / A2D / youtube and screen/pir submodule of A2D
defined tiner of screen/pir is for example : 2 min
duration of video +/- 4 min
A2D will turn on the monitor for duration of video (playing)
A2D will restart the counter at the end of video

Hey @Bugsounet
I am not sure I follow you very well … but the reason why I needed or used NEW-PIR was because it was the only thing that did not actually turn off the monitor or go to sleep so the the monitor power saver showed.
As far as I know NEW-PIR throws up a “black” screen or something so that the monitor never “sleeps” and when it wakes the monitor doesn’t report the input (hdmi) or show the power save on sleeping.

New-PIR was installed first… then I figured out how to install GA… then installed A2D, and then installed facial recognition …

It all seems to work fine the way I have it … honestly I couldn’t figure out the A2D commands and such in the beginning and do not want to reconfigure lol.
It all works great on my end for the moment I suppose.

majik655

@killerjoe34 said in MMM-GoogleAssistant with MMM-MotionDetector:

Hi guys
Thanks to @Bugsounet, GA and AD2 both work really well for me now with Spotify !
I installed MotionDetector wich works also well.

But when the display is off, I would like to turn it on by saying the hotword (Jarvis in my case)

How can I do that ?
Thanks

Re: MMM-GoogleAssistant with MMM-MotionDetector
I personally use MMM-NEW-PIR … and don’t need to wake the screen because the motion detector sees me and wakes the screen…this is using a PIR motion sensor though… so I am unsure you exact scenario…

But MAYBE you can look into the recipes…
In your assistant directory you will see a directory called RECIPES .
There are some premade recipes in there you can look over… you can also read about them in the wiki…
one of them is called “with-MMM-pages.js”

Open it and you will see the commands in that file that you can attach to your config file in the assistant module and it will do what the recipe tells it to…
I do not have access to this recipe file right now but I believe you can say “SHOW ALL MODULES” and the screen might wake up? (again this depends on your setup and your motion modules)…

You will need to add the recipe you want to use to your config file…
This page shows you the DETAILED config and at the bottom you will see RECIPES…
https://github.com/bugsounet/MMM-GoogleAssistant/wiki/Configuration-Sample

Add this to your config and put in the with-MMM-pages.js - restart Magic mirror and try out saying the commands.
You can also create your own recipes (look in the wiki) and add your own recipe js files to your config … you can use more than one.

    },
    recipes: [ "with-MMM-pages.js" ]
  }
},

majik655

@Blauenfeldt
install it without noobs. Noobs has given me problems in one way or another when I have used it. The issues range from a lot of things depending on what you are doing …

Go get the raspbian desktop image and just use that … it is just as easy to install as noobs and more than likely you will have working wifi when finished.

If you have another sd card (and dont want to erase your noobs one) just try it out on a second card… you will get it going.

majik655

@Feedy88 said in User Friendly Magic Mirror?:

@majik655 sorry, I didn’t read the original post properly, actually you are right.

@tripalreno4 actually I think @majik655 is right:

edit the wpa_supplicant.conf and add the gift receivers WiFi information

configure the MM to automatically start upon boot with pm2

This would basically make it a “plug and play” apart from configuring modules ofc.

No worries Feedy88 …

I have made a couple of these for people. From my own experience it depends on how many or what kind of modules you will be putting on the mirror.
Specialty modules that need their account info you will have to get for them and configure.

The more specialized you make it the harder it will be to “keep updated” so if you are going to keep it simple such as a clock, weather module and things like default modules then just getting the wifi up and running without them is probably the easiest along with using PM2 for auto boot … if mirror crashes or needs to turn off …pm2 will restart the mirror for them.

As for updating… you can turn off the update module notification (or leave it on I guess) but if you are using just the default modules there really isnt a reason to update unless something breaks.

majik655

Not sure if this will help you and sorry if this is obvious…
but if you want you could pre set up their wifi for them…

edit the wpa_supplicant.conf

When you connect to wifi the wpa_supplicant.conf file is created with your ssid’s and passwords etc etc…

So you can connect to your own wifi… then go edit the conf file and ADD their wifi …

When you first write the image to the sd card…there is no conf file because you havent connected to wifi yet (obviously)… so when done writing your sd card image… you can open the sd card (on widows or whatever) and create the wpa_supplicant.conf in the root with the info needed … otherwise just connect to wifi then edit the file.

Not sure if that helps you all that much but you can pre set it up you just need their info like SSID they would connect to and password

sudo nano /etc/wpa_supplicant/wpa_supplicant.conf

Below is what mine looks like with 2 ssid’s to choose from …

ctrl_interface=DIR=/var/run/wpa_supplicant GROUP=netdev
update_config=1
country=US

network={
        ssid="SSID HERE"
        psk="PASSWORDHERE"
        key_mgmt=WPA-PSK
}

network={
        ssid="SSID2 HERE"
        psk="PASSWORDHERE"
        key_mgmt=WPA-PSK
}

majik655

I can’t remember if turning on OpenGL driver was my issue or not with this…
As long as your config files is correct (it isnt a hard config file I will post mine below)…
Try opening a terminal window and type sudo raspi-config
go to “Advanced Options” and then “OpenGL Driver”

Depending on the hardware you have or what PI model you are using you may have just one option or a couple…
IF you have multiple options like I have seen … one was FAKE KMS and the other was FULL KMS…
I had to choose FULL KMS (i think for raspberry pi3)… on my raspberry pi 4 I believe there is only one option…choose it.

(might want to reboot)… cant remember if needed or makes you.

Then try again … make sure your device is on the same wifi network as your mirror …
You should see your mirror name …

I can’t think of what more there could be … unless your router is blocking the default port?
You did do npm install in the module directory after cloning it right?

My config for the module…

		{
			module: 'MMM-Screencast',
			position: 'center', // This position is for a hidden <div /> and not the screencast window
			config: {
				position: 'center',
				height: 300,
				width: 500,
			}
		},

majik655

@sdetweil
Yeah so one thing I have found is it is kind of worthless to have a motion sensor running that hides the modules after 2minutes of no motion…because if you block the sensor once it has closed… it looks like the camera cant turn it back on because the sensor has hidden everything.

I either turn off the motion sensor and rely on the camera all the time …
or
somehow have the motion detection turn on the camera recognition part…recognize then show correct modules.
or
have the recognition…trigger the motion sensor

so many decisions I dont know which to do lol

majik655

@majik655

Best posts made by majik655

Latest posts made by majik655