I am using those instructions: Training Custom Object Detector — TensorFlow 2 Object Detection API tutorial documentation in order to implement object detection on image. When object detection takes place, I get many green rectangles around objects. I need to find the image coordinates (x, y) of those rectangles. How can I achieve that?
Can you show the code where you use the model?
Probably there’s some cv2.rectangle call. There the rectangles are drawn to the image, so there you have access to them
The code is inside the above link. I do not use custom code…
the code in the link does not draw (or even detect) anything,
it’s all about training / building tf records.
please do not waste our time like this.
My friend, I followed the above instructions and did the right configurations, and ran the scripts. I did not write any code… The last thing I want is to waste your time. I am trying to do something here…
Please look here: python - Get the bounding box coordinates in the TensorFlow object detection API tutorial - Stack Overflow I use Vadim’s code. Moreover at the beginning of the site it says:
(left, right, top, bottom) = (xmin * im_width, xmax * im_width, ymin * im_height, ymax * im_height)
What do I put in
im ?? I cannot figure out something clear as image from the code…
Does anyone read the current post???
I get wrong coordinates!! I don’t know why…