RPi Cam Web Interface

[Gary Littlemore]

I think there is something wrong with update.sh script.
I launched it, commits didn't match so it updated, but i still see the old 6.2.28 version instead of the new v6.2.31. If i look at the installation directory i can see config.php timestamp changed to when i launched the update.sh script, but no modification of the APP_VERSION string.

So now i don't know if i have the last files from github or not.


Thank you btidey for your work!

p.s. the wiki, step 4, says ./upgrade.sh but i think should be ./update.sh

The update script from 6.2.25 uses a git fetch rather than a pull so should not be suffering from mismatch commits.

Your log shows that fetch is indeed being used and then runs the install script afterwards so that does look right. You can check in the RPi_Cam_Web_Interface and see if the files are updated. The config.php file in the www folder is where the version number is contained so you should see that near the top of the file.

I'll check further but one can also do the equivalent of an update just by deleting the RPi folder and doing the git clone followed by ./install.sh

I corrected the wiki reference. Thanks.

Any further with this issue, I'm still stuck with 6.3.1 not upgrading.

[IoT Maker DSM]

Example of automated video email?
Could a bash cowboy/cowgirl, please review my code below and provide guidance/corrections? I am following the instructions from elinux, but email notification section is sparse on details or an example.

I have the RPi Cam Web up and working. I installed, configured and verified that ssmtp and mpack are able to send emails/attachments from pi@raspberrypi prompt.

I need a second set of eyes or someone to verify this is will send a thumbnail once the video is written:
1. nano /var/www/macros/end_vid.sh
2. add:
!/bin/bash
list=( $1*.th.jpg )
thumb="${list[-1]}"
mpack -s "Motion Detected" /var/www/media/thumb mail.address@example.com
^---- is this right? thumb or should it be $thumb
[save and exit nano]

3. sudo chmod a+x /var/www/macros/end_vid.sh
4. sudo chown www-data:www-data /var/www/macros/end_vid.sh

Does this look correct? is there a more elegant way or method that I overlooked?

Guidance and help are appreciated! thank you!

[btidey]

Scheduler now has weekday checkboxes in the period command table. Commands are only issued for start stop and period change if the checkbox for the day is active.

By default all checkboxes are active so that behaviour is unchanged.

This facility is probably most useful when the Fixed TImes mode is used so the number of fixed time slots has been increased to 12.

[btidey]

Example of automated video email?
Could a bash cowboy/cowgirl, please review my code below and provide guidance/corrections? I am following the instructions from elinux, but email notification section is sparse on details or an example.

I have the RPi Cam Web up and working. I installed, configured and verified that ssmtp and mpack are able to send emails/attachments from pi@raspberrypi prompt.

I need a second set of eyes or someone to verify this is will send a thumbnail once the video is written:
1. nano /var/www/macros/end_vid.sh
2. add:
!/bin/bash
list=( $1*.th.jpg )
thumb="${list[-1]}"
mpack -s "Motion Detected" /var/www/media/thumb mail.address@example.com
^---- is this right? thumb or should it be $thumb
[save and exit nano]

3. sudo chmod a+x /var/www/macros/end_vid.sh
4. sudo chown www-data:www-data /var/www/macros/end_vid.sh

Does this look correct? is there a more elegant way or method that I overlooked?

Guidance and help are appreciated! thank you!

I think you are missing a # on !/bin/bash

That is good way to get thumbnail path from parameter. thumb is okay to set the thumb variable; further refs to it should be $thumb

The mpack bit looks ok except for using $thumb but it might be worth just testing that out directly on the command line by manually putting a thumbnail path in.

[rahlquist]

Two changes in latest version.

Web interface now puts up a confirm box when Reset settings is pressed.
A uconfig.bak file is created from current one before it is cleared.

Thank you for this! Sorry my first chance to get back to this thread since I posted.

Also when I have motion setting set for External (Isnt this the classic motion that was first implemented with this cam interface?) and click on Edit motion settings all I see is the 4 buttons, no settings. Is that to be expected?

[btidey]

Two changes in latest version.

Web interface now puts up a confirm box when Reset settings is pressed.
A uconfig.bak file is created from current one before it is cleared.

Thank you for this! Sorry my first chance to get back to this thread since I posted.

Also when I have motion setting set for External (Isnt this the classic motion that was first implemented with this cam interface?) and click on Edit motion settings all I see is the 4 buttons, no settings. Is that to be expected?

When external motion is used then it must be active in order to see the settings table. This is because the motion process must be running in order to respond to requests for its parameters. That is still working OK for me and I see a table of values.

If you are getting just 4 buttons when motion is running and no warning message then something must be wrong retrieving the parameters from motion. Check out trouble shooting section in wiki for tips on viewing motion values.

[Mechman]

Example of automated video email?
Could a bash cowboy/cowgirl, please review my code below and provide guidance/corrections? I am following the instructions from elinux, but email notification section is sparse on details or an example.

I have the RPi Cam Web up and working. I installed, configured and verified that ssmtp and mpack are able to send emails/attachments from pi@raspberrypi prompt.

I need a second set of eyes or someone to verify this is will send a thumbnail once the video is written:
1. nano /var/www/macros/end_vid.sh
2. add:
!/bin/bash
list=( $1*.th.jpg )
thumb="${list[-1]}"
mpack -s "Motion Detected" /var/www/media/thumb mail.address@example.com
^---- is this right? thumb or should it be $thumb
[save and exit nano]

3. sudo chmod a+x /var/www/macros/end_vid.sh
4. sudo chown www-data:www-data /var/www/macros/end_vid.sh

Does this look correct? is there a more elegant way or method that I overlooked?

Guidance and help are appreciated! thank you!

The thumb filename generation from the wiki is not working for me. I use this end_vid.sh:

Code: Select all

#!/bin/bash

thumb=$(echo $1 | cut -d '.' -f 1-2)*.th.jpg

echo 'Body text' | mail -a $thumb -r from@address -s "Subject" to@address

Additionally I installed postfix and heirloom-mailx to send the mails from the pi.

[btidey]

The thumb filename generation from the wiki is not working for me. I use this end_vid.sh:

Code: Select all

#!/bin/bash

thumb=$(echo $1 | cut -d '.' -f 1-2)*.th.jpg

echo 'Body text' | mail -a $thumb -r from@address -s "Subject" to@address

Additionally I installed postfix and heirloom-mailx to send the mails from the pi.

Strange. It works OK for me and if I do it manually from command line then it seems OK on standard Jessie. Do you know what sort of error or output you got?

[Mechman]

Strange. It works OK for me and if I do it manually from command line then it seems OK on standard Jessie. Do you know what sort of error or output you got?

If I call the wiki script with
./end_vid_wiki.sh /var/www/media/vi_0000_20170331_101056.mp4.h264

Code: Select all

#!/bin/bash
list=( $1*.th.jpg )
thumb="${list[-1]}"
echo $thumb

I get /var/www/media/vi_0000_20170331_101056.mp4.h264*.th.jpg

And my script
./end_vid_mech.sh /var/www/media/vi_0000_20170331_101056.mp4.h264

Code: Select all

#!/bin/bash
thumb=$(echo $1 | cut -d '.' -f 1-2)*.th.jpg
echo $thumb

returns /var/www/media/vi_0000_20170331_101056.mp4.v0000.th.jpg
which is the correct thumb file name.

I'm using the current Jessie.

[btidey]

Strange. It works OK for me and if I do it manually from command line then it seems OK on standard Jessie. Do you know what sort of error or output you got?

If I call the wiki script with
./end_vid_wiki.sh /var/www/media/vi_0000_20170331_101056.mp4.h264

Code: Select all

#!/bin/bash
list=( $1*.th.jpg )
thumb="${list[-1]}"
echo $thumb

I get /var/www/media/vi_0000_20170331_101056.mp4.h264*.th.jpg

And my script
./end_vid_mech.sh /var/www/media/vi_0000_20170331_101056.mp4.h264

Code: Select all

#!/bin/bash
thumb=$(echo $1 | cut -d '.' -f 1-2)*.th.jpg
echo $thumb

returns /var/www/media/vi_0000_20170331_101056.mp4.v0000.th.jpg
which is the correct thumb file name.

I'm using the current Jessie.

OK I think the issue here is you are using end_vid which runs immediately after the capture ends and so is passed the temporary captured data filename .mp4.h264 It can't then find the thumbnail as this doesn't have the temporary .h264 part in it.

The boxing process then converts this to .mp4 and deletes the original mp4.h264

The end_box script then is called with the final .mp4 name. That is the one I use normally and the name extraction then works OK. The end_vid script is intended primarily for use when automatic boxing is turned off and the script is doing something else.

The wiki referred to using end_vid for sending mails. I've changed that to end_box

[Mechman]

The wiki referred to using end_vid for sending mails. I've changed that to end_box

Thanks, that explains the issue!

[IoT Maker DSM]

All,
thank you very much for the feedback...Below is my script, named: "end_box.sh" in /var/www/macros. "end_box.sh" is owned by www-data:www-data and has a+x execute permissions.

A second set of eyes is always welcome as well as suggestions, but I believe it is working...I am a neophyte, so if anyone has suggestions on how to log an errors with this script, please contribute.
thanks again!

Code: Select all

#!/bin/bash
#
# Maximum file sizes for sendmail
MAXSIZE=25000000        # gmail max filesize
#MAXSIZE=10000000       # hotmail max filesize

# email recipients...replace <myname> with your username
RCVR1=<myname>@gmail.com
RCVR2=<myname>@yahoo.com

# get thumb filename
list=( $1*.th.jpg )
thumb="${list[-1]}"

# Get file size of original video
FILESIZE=$(stat -c%s "$1") 

#check file size
if [ $FILESIZE -gt $MAXSIZE ]
then  
#too big...send thumb
mpack -s "Motion Detected - video is too big to send" $thumb $RCVR1, $RCVR2
else
#just right...mailing file
mpack -s "Motion Detected" $1 $RCVR1, $RCVR2
fi

exit

[gordoncunningham]

To have a basic non-controllable view page (untested, but pretty sure it should work as stated):

1. When prompted for a directory during the RPi_Cam install, enter "admin"
2. Follow the instructions here to password protect the admin directory.
3. From the www directory, "sudo ln -T admin/cam_pic.php cam_pic.php"
4. Download the file from the post above and unzip it to www.

I did the testing for you, it works,
I took a slightly different approach, i put the two files in the document root , /var/www/ plus the link to cam_pic.php.
The owner and group of the file is www-data.
Is this safe this way ? Now i have to study PHP, to split up the index.php to get the pipan controls also in the public directory /var/www.
Thanks for the help
( found out how to reply with quotes , hope this works out this time)

I've followed this info and have had a little success in getting the initial RPCWI web page working with the camera view available without a password, and a password-protected subdirectory for the full page with admin capabilities, which is what I'd wanted. However, I've found that the full control page does not allow the various buttons to work nor can I edit the camera control parameters. I'm sure it's something simple, but I've checked permissions, ownership, paths, etc. and I can't find it. Note that I have years of experience with Linux, PCs, etc. but I just can't find the issue here on this RPi3. I'm using the 5MP RPi camera.

Also, I can't seem to find any indication of issues in any of the logs.

And one last thing outside of this issue - I see there are over 4000 posts in this forum topic thread. It makes searching and finding info very tedious. Is there another support forum for RPCWI? Or perhaps the owners of the forum might create a sub-forum for RPCWI so topics can be broken out better?

Thanks.

- Gordon

[DrMikeG]

This work is excellent but this thread is a pain.

I've done my best to read and search the 4000+ posts (with the numerous side tracks onto other projects).

What am I trying to do:
- Capture short time-lapse sequence (1-2 min) at highest possible quality, at the expense of frame rate.

From what I have understood:
1) This project seems to be the main (if not only) web previewer for pi cam
2) raspimjpeg is designed purely for jpeg/mpeg
3) Including raw layer data does not work for time lapse (only single capture)

What I intend to try:
Adding a new button the interface which when pressed will:
a) stop raspimjpeg
b) using raspistill to capture a fixed number of png images (hopefully these will be higher quality / uncompressed)
c) restart raspimjpeg to continue previewing

My questions are:
Q1) Does this sound practical?
Q2) Is there an easier way I have missed?

Many thanks for your time and expertise.

[quavoce]

The latest version now has a video split capability. Under camera settings there is a new setting called video split. If this is 0 then behaviour is as before with continuous recording between a video start and stop. If set non-zero then the recording will be stopped and restarted at this interval whilst vide is active. So if set to say 300 then a new recording will be made at 5 minute intervals. Each recording has its own thumbnail and behaves the same as if it had been manually stopped and restarted at this interval.

Thank you Robert for this useful feature - I'm hoping to set my Pi recording all day using this!

During normal motion detection, video recordings are started and motion detection continues to work by prolonging the recordings if movement still occurs. This indicates that motion detection is still possible while recording video. However, if I manually start recording video, the motion section button is greyed out. Is this intentional to prevent something from breaking?

My thoughts are to make this work like some commercial solutions by recoding video all day and logging the time to a text file when any motion is detected. I like the idea of having the whole day recorded just in case something happens outside my masked area. For example, my camera includes part of the road which I mask off to prevent motion triggering whenever cars drive past, but a couple of weeks ago there was an accident and it would have been useful to have a video recording of this event.

Fingers crossed that this is possible...

Thanks!

P.S. Is it possible to include the date/time stamp on videos when using the manual recording button please?

[frankieforceps]

The RPi Web interface is great! I've tried a number of options and this solution is by far the best.

I do have a quick question-
I'm trying to capture the mjpg stream in OpenCV. The VideoCapture constructor in OpenCV needs a web address "filename" to capture the stream. Such as [address of pi]/something.mjpg
It seems to be critical for the filename to end with ".mjpg"

What "filename" would I use in the case of the RPi Web Interface? I've inspected the web element and the selector for the video/image is #mjpeg_dest. The source is "cam_pic_new.php?....xxxxxxxxx".

Here is the code I'm using to open the video capture in OpenCV:
cv::VideoCapture cap;

cap.open("http://192.168.0.2/html/");


Can any of you guys help me with a proper filepath to capture this mjpg stream? I would be so grateful! I've been trying all sorts of combinations with no results!

[blitz79]

Below is my script, named: "end_box.sh" in /var/www/macros. "end_box.sh" is owned by www-data:www-data and has a+x execute permissions.

Are you sure your macros_path is /var/www/macros? It is /var/www/html/macros for me.

[ziolupo]

Dear all,
I have a couple of Raspberry with RPi Cam Control that, IMHO, is the best solution available.

I tried to install a new raspberry3 but I'm having issues.
If I install it in the default folder (/var/www/html) everything is working fine. But that folder is too crowded and I tried to install it in a subfolder (/var/www/html/rcam).
The problem is that in the subfolder is not working and I have always the "LOADING" message instead of the camera output.
THank you for your help!

ZioLupo

[audunbrekke]

Timelapse convert by shell script.
Hi. I am capturing timelapse images every dey from dusk to dawn. And I would like to Automaticaly make a timelapse video every day. I could run a script with crontab.
Is there a script or command to do the file rename and convertion to a timelapsevideo that I can run in shell ?

[Gavinmc42]

Totally new to this post, 163 pages , the longest running post?
Just got it running on a Pi3, apache, nginx did not work but lighttpd did after I did update/upgrade.

Works on full Raspbain, will try Jessie Lite later then see if I can get it going on PiCore.

In those 163 pages has anyone got it going on a Zero W?
Was thinking of a solar powered version so I don't have dig up the garden to run conduit and wires.

Got a Cluster hat so will it run on Zero's, with output via USB to a Pi2/3?

[btidey]

Totally new to this post, 163 pages , the longest running post?
Just got it running on a Pi3, apache, nginx did not work but lighttpd did after I did update/upgrade.

Works on full Raspbain, will try Jessie Lite later then see if I can get it going on PiCore.

In those 163 pages has anyone got it going on a Zero W?
Was thinking of a solar powered version so I don't have dig up the garden to run conduit and wires.

Got a Cluster hat so will it run on Zero's, with output via USB to a Pi2/3?

I haven't tried PiZeroW but I have multiple cameras running on PiZero both wirelessly and wired by using external usb adapters. I don't think there should be any issue with Zero W.

Jessie Lite also works (you may need to add git on first) and if you want to convert time lapse on the Pi itself then some additional gstreamer install is needed.

If you have any detail on why an apache didn't work then that would be interesting. Somebody else mentioned that but when I tried it on latest Jessie it all seemed OK.

[btidey]

Timelapse convert by shell script.
Hi. I am capturing timelapse images every dey from dusk to dawn. And I would like to Automaticaly make a timelapse video every day. I could run a script with crontab.
Is there a script or command to do the file rename and convertion to a timelapsevideo that I can run in shell ?

I don't have a specific script for this but a good starting point is the cmd template for initiating a convert via the browser. This uses the gstreamer process to do the conversion which is quite fast as it uses the GPU but you have to be careful to keep the image resolutions GPU compatible (e.g. 1920 x 1080 is OK but larger image sizes are not).

gst-launch-1.0 -v multifilesrc location=i_%05d.jpg caps=\"image/jpeg,framerate=\(fraction\)30/1\" ! jpegdec ! queue ! videoscale ! video/x-raw ! queue ! omxh264enc target-bitrate=15000000 control-rate=variable ! video/x-h264, profile=high ! h264parse ! mp4mux ! filesink location=

When this is used the preview.php first gathers all the files into a temp folder (e.g. t1005) renaming them to i_00000.jpg, i_00001.jpg etc. It then modifies the i_%05d.jpg to include the temp folder reference (t1005/i_%05d.jpg) and also appends the desired video filename on the end after location=

A script would need to do similar operations.

[btidey]

Dear all,
I have a couple of Raspberry with RPi Cam Control that, IMHO, is the best solution available.

I tried to install a new raspberry3 but I'm having issues.
If I install it in the default folder (/var/www/html) everything is working fine. But that folder is too crowded and I tried to install it in a subfolder (/var/www/html/rcam).
The problem is that in the subfolder is not working and I have always the "LOADING" message instead of the camera output.
THank you for your help!

ZioLupo

That should have been OK but I don't normally use multiple subfolders for an install. I'll check it out, but you can also use /var/www/rcam to avoid any overlap.

If you have other web sites in /var/www/html then it is also possible that part of the web server config needs to be adjusted to be compatible with those and the camera site.

[btidey]

The latest version now has a video split capability. Under camera settings there is a new setting called video split. If this is 0 then behaviour is as before with continuous recording between a video start and stop. If set non-zero then the recording will be stopped and restarted at this interval whilst vide is active. So if set to say 300 then a new recording will be made at 5 minute intervals. Each recording has its own thumbnail and behaves the same as if it had been manually stopped and restarted at this interval.

Thank you Robert for this useful feature - I'm hoping to set my Pi recording all day using this!

During normal motion detection, video recordings are started and motion detection continues to work by prolonging the recordings if movement still occurs. This indicates that motion detection is still possible while recording video. However, if I manually start recording video, the motion section button is greyed out. Is this intentional to prevent something from breaking?

My thoughts are to make this work like some commercial solutions by recoding video all day and logging the time to a text file when any motion is detected. I like the idea of having the whole day recorded just in case something happens outside my masked area. For example, my camera includes part of the road which I mask off to prevent motion triggering whenever cars drive past, but a couple of weeks ago there was an accident and it would have been useful to have a video recording of this event.

Fingers crossed that this is possible...

Thanks!

P.S. Is it possible to include the date/time stamp on videos when using the manual recording button please?

Sorry for delay in replying to this, busy with other stuff.

At the moment manual initiating a recording does grey out motion detection as these were thought to be mutually exclusive modes. The underlying process could have motion detection running even when doing manual recordings but it would need a change to do this. The action on motion detection in the scheduler could then be used to fire a script to log each motion detection.

[btidey]

And one last thing outside of this issue - I see there are over 4000 posts in this forum topic thread. It makes searching and finding info very tedious. Is there another support forum for RPCWI? Or perhaps the owners of the forum might create a sub-forum for RPCWI so topics can be broken out better?

Thanks.

- Gordon

I appreciate the difficulty finding material in this thread. I often have difficulties finding previous stuff.

One can raise issues for web side at https://github.com/silvanmelchior/RPi_C ... ace/issues
Similarly if issue is known to be with raspimjpeg process then https://github.com/roberttidey/userland/issues

It is good also to check the wiki http://elinux.org/RPi-Cam-Web-Interface before raising issues. It is good if any hints/ tips / clarifications are put in there to make it easier to find.