juangallostra / augmented-reality Goto Github PK

I want to be able to match more than a single reference object in the captured frames without any serious lags. I'm guessing threading is the way to go but still am unsure about how to go about it and match a reference object with the 3d model to be rendered.

Description

Code is written for Python 2.7, but it is quite straightforward to make it work with Python 3. Update the code so that the default Python version is Python 3.

Related PRs: #14 #11

Description

Instead of performing only detection on each individual frame, take into account the information from previous frames to track the reference surface and reduce jitter.

Approaches to test (or include):

• Mean filter
• Kalman filter

I tried loading a few models of my own but the loaded models are extremely tiny while models you have provided are of the perfect size. Is there any way to overcome this? Also, how do you go about loading textures?

Attaching a model that has this issue:
dna.zip

Hi,
When i run this code on a standard visiting card, it is not quite able to recognize it. The video is not input via webcam, but pre recorded and input using cv2.VideoCapture() . Is there a set dimension of image or something ?
Thanks!

When I change to my image as model.jpg, program error:

 File "src/ar_main.py", line 55, in main
    matches = bf.match(des_model, des_frame)
cv2.error: OpenCV(3.4.1) /Users/travis/build/skvark/opencv-python/opencv/modules/core/src/batch_distance.cpp:238: error: (-215) type == src2.type() && src1.cols == src2.cols && (type == 5 || type == 0) in function batchDistance

What should this error means?

Dear juangallostra:

I can show the AR model by the code ,however, the 3d model not stable（ flash of move）.
And a problem could I get the postion and rotate of the camera?
Thanks.

I am getting the error - Unable to capture video. Although, I have the opencv built with FFMPEG support as I checked my opencv build info. Will be great if you could help

I am getting this error any idea, please?

usage: ipykernel_launcher.py [-h] [-r] [-mk] [-fk] [-ma]
ipykernel_launcher.py: error: unrecognized arguments: C:\Users\Wooks\AppData\Roaming\jupyter\runtime\kernel-6769bc09-1f3c-4e3a-89c1-35e83a077bea.json
An exception has occurred, use %tb to see the full traceback.

SystemExit: 2

C:\ProgramData\Anaconda3\lib\site-packages\IPython\core\interactiveshell.py:2971: UserWarning: To exit: use 'exit', 'quit', or Ctrl-D.
warn("To exit: use 'exit', 'quit', or Ctrl-D.", stacklevel=1)

hi @juangallostra
when i ran your code, then error occured like below

[ INFO:0] Initialize OpenCL runtime...
Traceback (most recent call last):
File "src/ar_main.py", line 174, in
main()
File "src/ar_main.py", line 40, in main
obj = OBJ(os.path.join(dir_name, 'models/cow.obj'), swapyz=True)
File "/home/hyunjo/Face_Detection/augmented-reality/src/objloader_simple.py", line 16, in init
v = v[0], v[2], v[1]
TypeError: 'map' object is not subscriptable

can you tell me how to fix it ?
in advance, thank you

Currently the requirements file includes all the dependencies required to run both the AR app and the data visualization app. Furthermore, some of the data visualization dependencies are quite heavy (Flask, Plotly or Dash).

I think it makes sense to split the requirements in two parts (AR on one side and data visualization on the other side) so that users who only want to tinker with the AR code don't install packages that will not be used.

Additional info: Customize a requirements.txt for multiple environments

I'm trying to make it work through sockets but everything breaks down when I pass the base64 image

Hi,
The code is great. While running this the issue I am facing is the objects keep jumping around and do not stay on the reference surface. I am using the webcam. I am i doing something wrong. I am still not sure if the dimensions of the card or the surface of the card is the problem. Kindly let me know.

Dear juangallostra:

I can show the AR model by the code ,however, the 3d model not stable（ flash of move）.
And a problem could I get the postion and rotate of the camera?
Thanks.

thanks for your work!
Now i get a question:
Is a single color space covering the texture of the model? Or should we combine the MTL file to display the texture of the model？
thanks ！

Description

Currently only monochromatic models are supported. Add support for models with color.

Related issue: #10

how to use a colorful AR model?

This is a great job! I want to know if this job can be expanded to detect the ground and place a model on it instead of detecting cards. Thanks a lot!

matches = bf.match(des_model, des_frame)
          ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

cv2.error: OpenCV(4.9.0) D:\a\opencv-python\opencv-python\opencv\modules\core\src\batch_distance.cpp:275: error: (-215:Assertion failed) type == src2.type() && src1.cols == src2.cols && (type == CV_32F || type == CV_8U) in function 'cv::batchDistance'