First of all you need to check/ensure the camera's URL is correct and is working.
You should be able to drop the camera's URL into a browser and the feed should appear in the browser.
If it doesn't - then there is no way it will work with HSS.
Please check the above and get back to us with the result.
Hello @krambriw
Once again, I appreciate the work of you and your colleagues. I understand that there are so many cameras and protocols that it's difficult to attend them all.
In my case I have a camera with rstp and the video is an FFmpeg. Any tips on how to integrate it into your excellent node red flow.
( ps I got it work through vlc)
Hello Jose, thanks for the kind words, in fact, Paul, David and myself spent quite a lot of time building the HSS and we are very happy to share our joint project with the community
I think your camera should be possible to integrate without any major problems. I have described in some postings above how to add an ip camera like yours. Could you look into that? Also post the url (remove your user and password if you have that) you are using in vlc, that will help me to better understand and give you advice how to proceed.
Some ip cameras, I think like Axis, do provide streams also via http so those are dead simple to integrate. Is a pitty I do not have any ip cameras myself so I cannot test, that's why I have used and provided those public streams as example, some of them are actually from Axis cameras
for streaming I'm trying to use GitHub - arkhipenko/esp32-cam-mjpeg-multiclient: ESP32 MJPEG Multiclient Streaming Server with no success so far.
To check the camera out I've used the ESP32 camera webserver example and that does seem to work ok, so I know the cam does actually function correctly ( although it does take a few seconds for the video to appear when I first connect to it )
once I found the streaming address ( as opposed to the web page address ) and plugged that into HSS its all working ! seems to take a little while for the stream to appear, but then its fine
I have a few questions related to how I would go about transitioning from Sighthound Video to HSS. I haven't dug very deep yet.
Perhaps I'm mistaken, but given the feature list, I get the sense that HSS is geared toward a scenario where someone is sitting at a terminal (like a security guard) actively watching the feed, whereas I use Sighthound in a fully automated fashion, so let me ask a few questions about the features I use with SHV:
Can you create several rules per camera that, for example, let you:
Define an irregularly shaped region to watch for motion events
Detect motion from any object or a person
Define actions upon motion detection (e.g. export clip to a separate folder, sound an alarm, email a snapshot, call a webhook, etc.)
Enable/disable any of the rules
Manually (in the GUI)
Automatically (e.g. based on my Life360 node - I wrote a script to enable rules when the last one of my family leaves and disable when the first of us arrives home)
I have 6 cameras configured in SHV. Would they all be able to be monitored concurrently?
I had a setback trying to update using the Projects feature, and broke it. Not clear what I did wrong. THen something came up and I had to put things aside. Hopefully I can find some time to work on it a bit this weekend.
I'm going to wipe my flows file and start the installation over again. Then I'll try these instructions to get one or more of my IP camera rtsp streams playing. If I succeed here, I'm sure I'll have more questions eventually about modifying the flow to replace
You are welcome, will try to help as much as possible. Once you got the basics working I am particularly interested in re-designing the subflow part so you can use your external AI solution. As you know, I have such myself for my other video system in operation since a couple of years, using yolov4
Release version of the flow 'hss_complete_v1.1.0.zip' has been uploaded to the website.
This release includes the new version of Single File Gallery (SFG v4_8_0) which has many improvements and enhancements. We have packaged a new 'organise' script with this release as SFG needs to remove _sfpg_data as it now stores thumbnails in a different format.
You will need to re-run Step 1 and Step 2 on the HSS Setup SAS tab-flow and then all of this should happen automatically for you.
I hosed my node-red introducing the "projects" feature to update to v1.0.3. Not clear where or how I messed up, but now that I've recovered and have had the v1.0.3 project running overnight, I'll try the update to v1.1.0 before I attack adding my rtsp streams.
Where I'm a little unclear is how to I stop the current running project when I create and deploy the new one? Or does it happen automatically when I create/open the new project? I've never been motivated to try the projects feature so I didn't pay much attention to the discussions when it was introduced.
When we were first starting on our projects, you provided me some "rosetta stone" sample python code that really got me on the right track! I think it would be great if we can make this work.
I would strongly suggest switching on the 'projects' facility in Node-RED as it will make your life a lot easier when you come to upgrade/update to newer versions.
I did when I tried to update to v1.0.3 from v0.55.1 and managed to hose my node-red.
I've recovered and am running v1.0.3 as a project (the only one) now. My question is how do I stop the v1.0.3 project and start the new one? Or does it happen automatically when I create the new project? Can I rename a project" or do I clone it to a new name to accomplish it.
I haven't really thought thru the issues of starting to use the projects feature. It'd be nice to have project HSS be the current version and rename it to HSS_v1.0.3 when I update, but there could certainly be better ways. These development system "meta issues" can certainly be important in the long run, but they are initially a distraction when trying to just get something working.
No you can't rename a project at least what I know
But you can clone or you can actually export everything and then re-import it into a new project. Then delete the one with the name you do not like
To export everything, click on the first tab then go to the last tab at, hold shift and click on it. All tabs should now be selected, then select export, download to file. Open a new project with the name you like, then import what you just exported
I'm sure we will, already thinking that we if possible should modify/extend the current sub-flow in such way that we can have support for both the current internal coco and an external like yours. So on the input have some smart switching, selecting the path for the object detection. If triggered with some "external AI" attribute, sending the image data out (via MQTT) and then receieve the result including annotated image and detected filtered objects in some way. Sort of. Or if simpler, create a new parallel subflow. Not sure yet which one would be best and easy to implement
I just created a new project to update to HSS v1.1.0 and it couldn't have been smoother. Projects really needs a way to rename them as its hard to get a good naming scheme the first time you try as you are figuring it out.
Now that I've tried it, I'll probably always use it going forward. But it seems clunky at best, to take an existing flow and turn it into a project. Doubt I'll bother with it on my dedicated systems, but on my development system where I have many "disabled" flow tabs it'll clean things up nicely.
Next up, following your instructions to try and get some IP cameras in the Presets. About done for today unfortunately.