Docker instructions just exit without effect about camodet

[oyzzo]

Hi @adrianmay,

The camodet project has been unmaintained for a long time, it could be some versions issue.

There is a new python rewrite of the tool that you can use easily:
https://github.com/jim17/camodet_python/releases
https://github.com/jim17/camodet_python

Hope it helps!
I'll update the README and point to the new python version.

[adrianmay]

Thanks Noel. Just checking it out. I just bounced off zoneminder cos I couldn't build/configure my own code changes, and this looks like a good place to start. My goal is to have a motion detection alg that computes differences of pixel colours from their long term typical values, then tracks the centre of gravity of the set of active pixels. If I can distinguish changes towards vs away from the long term norm, then I can pick out both the leading and the trailing edge of a moving object. Things approaching the camera directly also need detecting. By "typical" I don't just mean average. I'd like to cater for e.g. green leaves waving against a blue sky and specifically recognise the two colours and mixtures of them (I'll whack the resolution down at the outset). Does that sound like fun? Adrian.

…

On Sun, 26 Apr 2020 at 09:36, Noel Carriquí Ojeda ***@***.***> wrote: Hi @adrianmay <https://github.com/adrianmay>, The camodet project has been unmaintained for a long time, it could be some versions issue. There is a new python rewrite of the tool that you can use easily: https://github.com/jim17/camodet_python/releases https://github.com/jim17/camodet_python Hope it helps! I'll update the README and point to the new python version. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#6 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAARMRZ3R56PNXJWGHB6I2DROPXBZANCNFSM4MRD3QUA> .

[oyzzo]

Hi @adrianmay,
It sounds great! It's easy to use this code or the python version as a start point and add your changes.
Using OpenCV is really easy to implement your described algorithm.

I implemented this code because zoneminder (and the other alternatives) was too complicate to install and setup, having a simple tool that does exacly what you need is great.

Using the C++ version requires to prepare the dependencies either by compiling them or install them which can be a time consuming task depending on the development platform.
Using the python version is as easy as "pip install" the needed libraries in any platform :)

Good luck with your project!

[adrianmay]

Thanks. I'll keep it on GitHub so you can pull the alg if you like it.

…

On Sun, 26 Apr 2020, 11:26 Noel Carriquí Ojeda, ***@***.***> wrote: Hi @adrianmay <https://github.com/adrianmay>, It sounds great! It's easy to use this code or the python version as a start point and add your changes. Using OpenCV is really easy to implement your described algorithm. I implemented this code because zoneminder (and the other alternatives) was too complicate to install and setup, having a simple tool that does exacly what you need is great. Using the C++ version requires to prepare the dependencies either by compiling them or install them which can be a time consuming task depending on the development platform. Using the python version is as easy as "pip install" the needed libraries in any platform :) Good luck with your project! — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#6 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAARMRYJYDNRSMT7OCM4DADROQD6RANCNFSM4MRD3QUA> .

[adrianmay]

Hi Noel, I'm getting a strange effect where the frames in the video come out in the wrong order. The go 1 2 5 3 4 6 etc. In the input view window they look fine. I can see this cos my ip cctv only seems to manage 1 fps. Did you ever notice that? I was also gobsmacked that the waitKey seems necessary to make the input view window show any frames. Was the C version this temperamental? Adrian. On Sun, 26 Apr 2020 at 15:32, Adrian May <[email protected]> wrote:

…

Thanks. I'll keep it on GitHub so you can pull the alg if you like it. On Sun, 26 Apr 2020, 11:26 Noel Carriquí Ojeda, ***@***.***> wrote: > Hi @adrianmay <https://github.com/adrianmay>, > It sounds great! It's easy to use this code or the python version as a > start point and add your changes. > Using OpenCV is really easy to implement your described algorithm. > > I implemented this code because zoneminder (and the other alternatives) > was too complicate to install and setup, having a simple tool that does > exacly what you need is great. > > Using the C++ version requires to prepare the dependencies either by > compiling them or install them which can be a time consuming task depending > on the development platform. > Using the python version is as easy as "pip install" the needed libraries > in any platform :) > > Good luck with your project! > > — > You are receiving this because you were mentioned. > Reply to this email directly, view it on GitHub > <#6 (comment)>, or > unsubscribe > <https://github.com/notifications/unsubscribe-auth/AAARMRYJYDNRSMT7OCM4DADROQD6RANCNFSM4MRD3QUA> > . >