search

cozmo / jsQR

A pure javascript QR code reading library. This library takes in raw images and will locate, extract and parse any QR code found within.
https://cozmo.github.io/jsQR/
Apache License 2.0
3.67k stars 606 forks source link

I cannot decode image of front mobile #126

Closed jonathasfujii closed 5 years ago

jonathasfujii commented 5 years ago

I cannot decode image of front mobile.

atkulp commented 5 years ago

Can you provide any more details? Not a developer of it, but I've tested it with great results. You need a camera with good enough quality of course though it's pretty forgiving.

jonathasfujii commented 5 years ago

The library works very well with the rear camera of the phone, but when using the front camera can not recognize the qr-code, the equipment used is a Smasung S7.

Atenciosamente,

Jonathas Fujii

Em qui, 21 de mar de 2019 às 02:00, Arian Kulp notifications@github.com escreveu:

Can you provide any more details? Not a developer of it, but I've tested it with great results. You need a camera with good enough quality of course though it's pretty forgiving.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/cozmo/jsQR/issues/126#issuecomment-475120872, or mute the thread https://github.com/notifications/unsubscribe-auth/ADkOUREo4Hv_fMX5fQCjDdszPq4GGOpoks5vYyAUgaJpZM4cAJcJ .

atkulp commented 5 years ago

As I understand it the camera makes no difference. The library is just looking at an image, in your case, the current frame of the video stream. Perhaps your cameras have different resolutions and what you're passing isn't lined up right. Since the library doesn't do the actual frame capture, make sure you have a preview window so you can see that it's capturing properly, and make sure your code isn't make any assumptions about the canvas image.

cozmo commented 5 years ago

@atkulp is correct in that this library just processes image data and as such any issues should be results of aspects of the image, not the device that captured it. As such going to close this issue but pleas feel free to comment or re-open with specific failure cases that should succeed.