set up

ThasnimolVSam22d007 · 2023-02-20T08:23:10+00:00

we bought an oak d pro poe camera, for the gaze estimation application, so once we have the camera, what all steps need to do from starting, mainly need imag...

erik

Hi ThasnimolVSam22d007 ,
Please use the M12 to RJ45 ethernet cable you got with your camera. The plastic cap is for M8 connector. Please see docs here.
Thanks, Erik

ThasnimolVSam22d007

Can you please give online session for how to conenct and capture images, means how to see the images in that above demo, what all softwares needs install and all

erik

Hi ThasnimolVSam22d007 ,
For setup, please follow docs here.
Thanks, Erik

ThasnimolVSam22d007

erik
OK SIR

ThasnimolVSam22d007

mEANS I WANT TO KNOW , FOR WORKING WITH THIS CAMERA, WHICH ALL THINGS NEED TO INSTALL,?
ALSO, BASICALLY HOW TO CAPTURE IMAGES USING THAT SOFTWARE, THEN WILL TAKECARE OF REMAINING THINGS, , CAN YOU PLEASE HELP ME TO SORT IT OUT?

ThasnimolVSam22d007

one more thing i connected poe injector to the system , to the camera, and one pin to power supply,
so apart from this one extra wire and one programming wire is with the kit,so where to connect those two? (image of both is attached)

ThasnimolVSam22d007

What is the use of that programming board, ididnt see anything related to it in documentation

ThasnimolVSam22d007

Basically i want to check whether the camera which i bought is working or not? initially? how to do this
currently, when am connecting the camera to poe of poe injector, then its lan to the system, but system only having one port, so how to match the ip address for all? i am using neutrik poe, so i how come i will match all ip adress while using this device

ThasnimolVSam22d007

Hi erik,
thank you for your help
i completed the installation succesfully, but while running demo, ir led s are not turning on? how to make it turning on ?
i want to capture some monocamera images using i, how to do it

erik

Hi ThasnimolVSam22d007 ,
Please see docs here.
Thanks, Erik

ThasnimolVSam22d007

hi erik, without saving images into the system , here we can use mobile nn, so i found some example with person object detection, so is there any models for detecting face ??

if any examples are there, kindle please share me, i need to do gaze detection without svaing images into the system

jakaskerl

Hi ThasnimolVSam22d007
Example for gaze detection.

Hope this helps,
Jaka

ThasnimolVSam22d007

jakaskerl

i saw this sir, but this is by saving video or directly?? bcausae when i pass arguments it is showing the frame, but how come i know which kind of results it is saving instaed of showing video to me?

jakaskerl

Hi ThasnimolVSam22d007
This example works by running gaze estimation on either live camera feed or pre-saved video on your machine. Either way, nothing gets saved back to your machine. It is only used to display the results.

-cam will use your camera and run neural network models on the live video;
-vid will load a saved video file and use it instead of live video from camera

Hope this helps,
Jaka

ThasnimolVSam22d007

jakaskerl

Hi,

i have one more query, this camera is actually for edge computing knw (means need to do all processing in camera itself), i am using gaze estimation code and face detection codes, but how can i verify whether that code processing is happening in camera ? not consuming cpu??

erik

Hi ThasnimolVSam22d007 ,
You can just check in the Task Manager how much the python script consumes CPU/GPU. All the AI happens on the camera itself for this demo. Note that it sends the results back to the host computer for decoding, and for that I already started working on PR here that also does decoding+cropping on the device (via script node). Thoughts?
Thanks, Erik

ThasnimolVSam22d007

erik

the pr link you given know? i didnt get any data from that page
with that code i can only preview , i tried to save video but not saving bbox in it, so if i want to save result only to the file, how to do it?? is it possible
by using openvino mobile ssd models only detecting person, and some other objects, is it possible to detect face,shoes, goggles and all using it? if you have any ideas regarding this just let me know

ThasnimolVSam22d007

erik

i tried this today, but that code with preview video is fully consuming cpu, after running the code consumption is incresed to 30,35 and all, so i am trying to save the results, do you have any ideas for it?? how to take output

erik

Hi ThasnimolVSam22d007 ,

It's just a work in progress PR.
Yes, you can save only the video. See DepthAI SDK's Recording module: https://docs.luxonis.com/projects/sdk/en/latest/features/recording/
You'd need to check the objects the pretrained mobilenet-ssd was trained on - I don't think it detects shoes and goggles. You could find a different pretrained model, or train your own neural network.

ThasnimolVSam22d007

erik

no i was asked, what is the purpose that link i couldnt find anything there
okay images am saving actually, in that color camera code but size is set to 300, when am trying to change it it is showing error only not changing the size? is possible to change the size of image or fixed??i mean is it possible to increase the field of view of image
from where i should get the objects pretrained in that network obkjects from that model, if you have any links please share here, am checking it also

ThasnimolVSam22d007

jakaskerl

means in that face detection code in github , using one model for object detection, so i want to know what all models are already defined in it, if it is not i want to include those things in the model, whether it is possible to add it in the same model,is it possible do you know whether anyone tried this, or any pretrained models with

shoes, goggles, person, face and all already available??
also erik i tried

that recording code it is saving some .dat file what is the advantage of those files?/ where i can use that??

jakaskerl

Hi ThasnimolVSam22d007

Its a link to a pull request which has the code that aims to do decoding/cropping via script node.
Your preview img size is (usually) set to whatever your NN expects. You can instead save video/still/isp to get a larger frame size. Info here. Basically have one stream (preview) go through NN, and the second (video) go directly through xlink/encoder to a file.
i'm sorry, what object are you referring to?

Thanks,
Jaka