[tool] [tutorial] Free AI Person Detection for Blue Iris

balucanb

Getting the hang of it
Joined
Sep 19, 2020
Messages
141
Reaction score
23
Location
TX
I have never needed additional deepstack performance but looking at your setting you are only pointing to one machine and using two ports. You will need to give aitool the address of the second machine running DS i.e 192.168.1.10: XX
Thanks @Village Guy That is what I thought, However I do see it in the log, it was a error but something is seeing the address/port? Now mind you I have no idea what I am looking at really- So going back to your suggestion- I know this is a stupid question but how do I get the address you are talking about? Update- so I went into history and re tested a image that I knew was good and it errored out on that port (82) so I think you are correct
 

Attachments

Village Guy

Pulling my weight
Joined
May 6, 2020
Messages
223
Reaction score
100
Location
UK
Thanks @Village Guy That is what I thought, However I do see it in the log, it was a error but something is seeing the address/port? Now mind you I have no idea what I am looking at really- So going back to your suggestion- I know this is a stupid question but how do I get the address you are talking about? Update- so I went into history and re tested a image that I knew was good and it errored out on that port (82) so I think you are correct
The machine that has the second deepstack installation will have it's own IP address. That is the address you will use along with the port you have assigned in the second deepstack installation. it can be the same port address as used for the first install.

i.e 127.0.0.1:81, 192.168.1.10:81

On the second machine you can find the IP address under: Control Panel\Network and Internet\Network Connections
 

balucanb

Getting the hang of it
Joined
Sep 19, 2020
Messages
141
Reaction score
23
Location
TX
The machine that has the second deepstack installation will have it's own IP address. That is the address you will use along with the port you have assigned in the second deepstack installation. it can be the same port address as used for the first install.

i.e 127.0.0.1:81, 192.168.1.10:81

On the second machine you can find the IP address under: Control Panel\Network and Internet\Network Connections
Yes Sir! I was being lazy, I'm blaming the turkey coma...did a 2 sec Google search and figured it out. It is all working now, I don't think I really need a second instance I was just curious.
 

balucanb

Getting the hang of it
Joined
Sep 19, 2020
Messages
141
Reaction score
23
Location
TX
So my experiment to see if I could get another instance of Deepstack working did not work- I don't need another instance so it is not a show stopper but I am curious what I did wrong so any help is appreciated. I am running the Windows version and everything is loaded on this computer- (BI, AI-Tool, Deepstack) I loaded a new version of Deepstack onto another Windows computer, I tested everything as far as making sure it was activated. I used the IPV4 address of the new computer and tried adding a different port then I was using on the main machine, it errored out then I changed it to the same port I was using on the main machine, still errored out (see below for error) The only thing that is different between the 2 computers is the one that works is Win 10 Pro and the 2nd one is just a Win10 Home. When I tried to add the 2nd instance in the settings screen in AI TOOL I used a comma to separate them and I also tried " ; " same result either way. the Here are the errors with the different ports-

11/27/2020 3:01:14 AM DetectObjects Unable to connect to the remote server [WebException] Mod: <GetDetectionsFromAIServer>d__35 Line:992:21 Error AITOOLS.EXE 192.168.0.23:82 RSyardsd RSyardsd.20201127_030046280.jpg 1853 1 60 True aitool.[2020-11-27].log
11/27/2020 3:01:14 AM ImageQueueLoop ...Problem with AI URL: '' (URL ErrCount=4, max allowed of 5) Info AITOOLS.EXE 192.168.0.23:82 RSyardsd 1859 0 60 True aitool.[2020-11-27].log
11/27/2020 3:01:14 AM ImageQueueLoop ...Putting image back in queue due to URL '' problem (QueueTime=0.4, Image ErrCount=0, Image RetryCount=1, URL ErrCount=4): 'D:\AIinput\RSyardsd.20201127_030046280.jpg', ImageProcessQueue.Count=0 Info AITOOLS.EXE 192.168.0.23:82

Here is when I changed it to port 80-

11/27/2020 1:17:18 PM DetectObjects Unable to connect to the remote server [WebException] Mod: <GetDetectionsFromAIServer>d__35 Line:992:21 Error AITOOLS.EXE 192.168.0.23:80 LSyardsd LSyardsd.20201127_131657043.jpg 922 1 57 True aitool.[2020-11-27].log
11/27/2020 1:17:18 PM ImageQueueLoop ...Problem with AI URL: '' (URL ErrCount=4, max allowed of 5) Info AITOOLS.EXE 192.168.0.23:80 LSyardsd 928 0 57 True aitool.[2020-11-27].log
11/27/2020 1:17:18 PM ImageQueueLoop ...Putting image back in queue due to URL '' problem (QueueTime=0.4, Image ErrCount=0, Image RetryCount=1, URL ErrCount=4): 'D:\AIinput\LSyardsd.20201127_131657043.jpg', ImageProcessQueue.Count=0 Info AITOOLS.EXE 192.168.0.23:80 LSyardsd LSyardsd.20201127_131657043.jpg 929 0 57 True aitool.[2020-11-27].log
11/27/2020 1:21:58 PM WaitForNextURL ---- Trying previously failed URL again after 301 seconds. (ErrCount=4, Setting 'MinSecondsBetweenFailedURLRetry'=30): Info AITOOLS.EXE 127.0.0.1:80 LSyardsd 948 0 33 True aitool.[2020-11-27].log

What did I miss? THX.
 

joshwah

Pulling my weight
Joined
Apr 25, 2019
Messages
224
Reaction score
104
Location
australia
Is anyone running the latest version.. 1.0.57.7637? For some reason it just sits on "processing image..." and everything queues up... i've tried rebuilding, rebooting, etc.
 

balucanb

Getting the hang of it
Joined
Sep 19, 2020
Messages
141
Reaction score
23
Location
TX
Is anyone running the latest version.. 1.0.57.7637? For some reason it just sits on "processing image..." and everything queues up... i've tried rebuilding, rebooting, etc.
which fork are you using? A few people (not me-do not know how) have compiled new versions I'm using one that says 2.0.109.7617 the latest one I am aware of from VorlonCD/bi-aidetection, (Chris Dodge) is 2.0.37.7608
 

Tinbum

Pulling my weight
Joined
Sep 5, 2017
Messages
439
Reaction score
118
Location
UK
 

whoami ™

Pulling my weight
Joined
Aug 4, 2019
Messages
150
Reaction score
121
Location
South Florida
Does anyone know why the Simple motion detection algorithm is suggested over Edge Vector other than to save CPU? Or is to conserve resources the only reason?
 

whoami ™

Pulling my weight
Joined
Aug 4, 2019
Messages
150
Reaction score
121
Location
South Florida
I'm pretty sure for Edge Vector to work you have to have "Use Zones and Hot Spots" enabled using at least "Zone A" with the whole screen selected.

You can also test your settings and see how motion detection is working by selecting a video to test by playing it back in Blue Iris directly in the console and right clicking and selecting "Test Run Video Through Motion Detector"
 

Village Guy

Pulling my weight
Joined
May 6, 2020
Messages
223
Reaction score
100
Location
UK
I'm pretty sure for Edge Vector to work you have to have "Use Zones and Hot Spots" enabled using at least "Zone A" with the whole screen selected.

You can also test your settings and see how motion detection is working by selecting a video to test by playing it back in Blue Iris directly in the console and right clicking and selecting "Test Run Video Through Motion Detector"
I guess at the end of the day you choose what works best for your application. I spent three days testing in a real world application and found Edge Vector detection failed miserably when I walked around the periphery of the field of view. Edge Vector definitely generates less false alarm but in my experience at the expense of missing real periphery alarm events.

FYI If you are not already aware, simple mode was the default for version 4.
 
Top