Extract images object of Tensorflow object detection API

Question

My problem is that I have ymin,xmin,ymax,xmax and boxes shapes.

I can't extract objects detected, it just display all image not boxes.

What I tried:


image_url = "https://upload.wikimedia.org/wikipedia/commons/6/60/Naxos_Taverna.jpg"  #@param
downloaded_image_path = download_and_resize_image(image_url, 1280, 856, True)

import cv2

...

 for i in range(min(boxes.shape[0], max_boxes)):
    if scores[i] >= min_score:
      ymin, xmin, ymax, xmax = tuple(boxes[i])
      display_str = "{}: {}%".format(class_names[i].decode("ascii"),
                                     int(100 * scores[i]))
      color = colors[hash(class_names[i]) % len(colors)]
      image_pil = Image.fromarray(np.uint8(image)).convert("RGB")
      print(image_pil)
      image = np.array(image_pil)
      print(type(image))
      import cv2
      rect = cv2.rectangle(image, ((xmin), (ymin)), ((xmax), (ymax)), (0, 0, 255), 1)

      print(ymin, xmin, ymax, xmax)
      cv2_imshow(rect)

I use this API: https://www.tensorflow.org/hub/tutorials/object_detection

Example of one object detection ymin, xmin, ymax, xmax : 0.6325336 0.2925815 0.92287964 0.40271035 object type : b'Chair'

Can you please add a sample image with sample coordinates (ymin, xmin, ymax, xmax) to test your code with? minimal reproducible example. — Bilal
– Bilal, Commented Oct 18, 2020 at 8:27
ymin, xmin, ymax, xmax : 0.6325336 0.2925815 0.92287964 0.40271035 object type : b'Chair' image_url = "upload.wikimedia.org/wikipedia/commons/6/60/Naxos_Taverna.jpg" #@param — mrNicky
– mrNicky, Commented Oct 18, 2020 at 10:08

Bilal · Accepted Answer · 2020-10-19 03:04:44Z

1

The Coordinates of the Rectangle has to be integers, not floats!

~~ymin, xmin, ymax, xmax : 0.6325336 0.2925815 0.92287964 0.40271035~~
Another issue is the start point, and endpoint coordinates should be in this format

(xmin, ymin), (xmax, ymax), adding unnecessary parentheses((),()) will define integers as tuples, and that will cause errors.

xmin = int(0.2925815 * img.shape[1])
xmax = int(0.40271035 * img.shape[1])
ymin = int(0.6325336 * img.shape[0])
ymax = int(0.92287964 * img.shape[0])

rect = cv2.rectangle(img, (xmin, ymin), (xmax, ymax), (0, 0, 255), 1)

answered Oct 19, 2020 at 3:04

Bilal

4,1256 gold badges33 silver badges70 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

mrNicky Over a year ago

I know, thanks it works! And then I just use image[ymin:ymax, xmin:xmax]

Collectives™ on Stack Overflow

Extract images object of Tensorflow object detection API

1 Answer 1

1 Comment

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Related