huningxin / opencv.js Goto Github PK

I am trying to write markerless ar on js by using open CV.
I am very early in JS (I am C++/python developer).
In standart opencv.js package I don't find this Structures

• Point3f
• solvePnP
• findHomogpaphy
• Аnd etc.

As I understand I need to compile some opencv modules (features 2d, calib3d, impoc, objdetect, ml) to js. How to do it? Do you have pre-compiled modules?

Could you add license to this project? I'm working on something and using your wasm build, is it ok if I reupload it somewhere?

I have embinded the function 'readNetFromModelOptimizer' into opencv.js and try it to load OpenVINO model in browser, then the following error occured:
error: (-2:Unspecified error) Build OpenCV with Inference Engine to enable loading models from Model Optimizer. in function 'readFromModelOptimizer' "

It seems like that we need to initialize OpenVINO environment in browser before using opencv.js to load OpenVINO model.

Hi !
Since Wasm permit to get library like OpenCV in browser, I would like to known if you think it's possible to use it to generate facial landmarks on the client side ?
This could be then used as an identification system on websites, without the need to send theses informations to the server ( just a salt/verifier SRP stored on the server ).
Regards,

Hi @huningxin hu

I am trying to do some experiments with Opencv.js and web worker. What I did clone OpenCV to my machine, run docker to build opencv.js as official docs tutorials (https://docs.opencv.org/3.4.13/d4/da1/tutorial_js_setup.html). But when I tried to copy the file opencv.js and loaded it in the worker.js file by importScripts, it failed. Do you have any experience with this problem? If yes, I really appreciate it if you could let me know what I am doing wrong.

Thank you 🙏