Embedding Laplacian variance inside Tensorflow bounding box [webcam]

Asked May 24 '19 at 10:05

Active May 24 '19 at 10:24

Viewed 102 times

I am trying to make bounding box which will also show focus quality prediction like in example shown here:

Image from here

I want to do it in real time, using webcam.

I am already using visualization_utils to draw a boxes on my output (camera, opencv). Later I use function cv2.Laplacian(image, cv2.CV_64F).var(), but it gives me overall value of whole picture.

My goal is to calculate Laplacian variation inside the bounding box.

Could you please give me an advice or provide an example how it can be done?

edited May 24 '19 at 10:24

Adam

2,726
1
9
22

asked May 24 '19 at 10:05

stacker

try.. https://stackoverflow.com/questions/48108717/opencv-laplacian-function-not-work-effectively-like-in-document – VeNoMouS May 24 '19 at 10:12
so basically "how to compute the laplacian only inside a rectangle"? – Miki May 24 '19 at 10:18
@Miki Yes, but using visualization_utils I am not sure if it is possible to do :( – stacker May 24 '19 at 10:21
Thats what i thought, i thought one would dupe the image where rect is and crop... and do laplacian .. but i dunno... – VeNoMouS May 24 '19 at 10:21
There is a function `tf.image.crop_to_bounding_box` but I am not sure how to use it correctly – stacker May 24 '19 at 14:35

Embedding Laplacian variance inside Tensorflow bounding box [webcam]

0 Answers0