20

I'm trying to return list of objects that have been found at image with TF Object Detection API.

To do that I'm using print([category_index.get(i) for i in classes[0]]) to print list of objects that have been found or print(num_detections) to display number of found objects, but in both cases it gives me list with 300 values or simply value [300.] correspondingly.

How it`s possible to return only that objects that are on image? Or if there is some mistake please help to figure out what is wrong.

I was using Faster RCNN models config file and checkpoints while training. Be sure it really detects few objects at image, here it is:

enter image description here

My code:

import numpy as np
import os
import six.moves.urllib as urllib
import sys
import tarfile
import tensorflow as tf
import zipfile

from collections import defaultdict
from io import StringIO
from matplotlib import pyplot as plt
from PIL import Image

from object_detection.utils import label_map_util
from object_detection.utils import visualization_utils as vis_util

PATH_TO_CKPT = 'frozen_graph/frozen_inference_graph.pb'

PATH_TO_LABELS = 'object_detection/pascal_label_map.pbtxt'

NUM_CLASSES = 7

detection_graph = tf.Graph()
with detection_graph.as_default():
  od_graph_def = tf.GraphDef()
  with tf.gfile.GFile(PATH_TO_CKPT, 'rb') as fid:
    serialized_graph = fid.read()
    od_graph_def.ParseFromString(serialized_graph)
    tf.import_graph_def(od_graph_def, name='')

label_map = label_map_util.load_labelmap(PATH_TO_LABELS)
categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)
category_index = label_map_util.create_category_index(categories)

def load_image_into_numpy_array(image):
  (im_width, im_height) = image.size
  return np.array(image.getdata()).reshape(
      (im_height, im_width, 3)).astype(np.uint8)


PATH_TO_TEST_IMAGES_DIR = 'object_detection/test_images/'
TEST_IMAGE_PATHS = [ os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image{}.jpg'.format(i)) for i in range(1, 2) ]

IMAGE_SIZE = (12, 8)

with detection_graph.as_default():
  with tf.Session(graph=detection_graph) as sess:
    sess.run(tf.global_variables_initializer())
    img = 1
    for image_path in TEST_IMAGE_PATHS:
      image = Image.open(image_path)
      image_np = load_image_into_numpy_array(image)
      # Expand dimensions since the model expects images to have shape: [1, None, None, 3]
      image_np_expanded = np.expand_dims(image_np, axis=0)
      image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
      # Each box represents a part of the image where a particular object was detected.
      boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
      scores = detection_graph.get_tensor_by_name('detection_scores:0')
      classes = detection_graph.get_tensor_by_name('detection_classes:0')
      num_detections = detection_graph.get_tensor_by_name('num_detections:0')

      (boxes, scores, classes, num_detections) = sess.run(
          [boxes, scores, classes, num_detections],
          feed_dict={image_tensor: image_np_expanded})

      vis_util.visualize_boxes_and_labels_on_image_array(
          image_np,
          np.squeeze(boxes),
          np.squeeze(classes).astype(np.int32),
          np.squeeze(scores),
          category_index,
          use_normalized_coordinates=True,
          line_thickness=8)
      plt.figure(figsize=IMAGE_SIZE)
      plt.imsave('RESULTS/' + str(img) + '.jpg', image_np)
      img += 1

      # Return found objects
      print([category_index.get(i) for i in classes[0]])
      print(boxes.shape)
      print(num_detections)

Which gives following result:

[{'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'chesterfield_blue', 'id': 1}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_gold', 'id': 5}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_red', 'id': 7}, {'name': 'chesterfield_red', 'id': 2}, {'name': 'marlboro_red', 'id': 7}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_red', 'id': 7}, {'name': 'lucky_strike_blue', 'id': 3}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'lucky_strike_red', 'id': 4}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'marlboro_mentol', 'id': 6}, {'name': 'lucky_strike_red', 'id': 4}]
(1, 300, 4)
[ 300.]

Thanks in advance for any information!

UPD:

Thousand thanks for everyone who helped with this question. Following line of code is exactly what I needed, it gives me list with objects that were found so I can do other operations on them.

print [category_index.get(value) for index,value in enumerate(classes[0]) if scores[0,index] > 0.5]
Improve this question

6 Answers 6

Reset to default
10
+50

As far as I can see you have 300 detections. visualize_boxes_and_labels_on_image_array shows very few of them because min_score_thresh=.5 (this is the default value) is too high for the most of them.

If you want to add such filtering to the output you can write:

min_score_thresh = 0.5
print([category_index.get(i) for i in classes[0] if scores[0, i] > min_score_thresh)

You can change min_score_thresh to choose threshold value you need. It may be useful to print the score values with the category names.

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

2 Comments

oh, I think thats it, I caught the main idea... I will try to figure out a little bit later, but first try gave me this error IndexError: only integers, slices (':'), ellipsis ('...'), numpy.newaxis ('None') and integer or boolean arrays are valid indices, did u try your code?
ok, I have modified just a little bit your code, and following statement gives me exactly what I needed print [category_index.get(value) for index,value in enumerate(classes[0]) if scores[0,index] > 0.5]
5

From the function signature visualize_boxes_and_labels_on_image_array, you have to set the arguments max_boxes_to_draw, min_score_thresh,

visualize_boxes_and_labels_on_image_array(image,
                                          boxes,
                                          classes,
                                          scores,
                                          category_index,
                                          instance_masks=None,
                                          keypoints=None,
                                          use_normalized_coordinates=False,
                                          max_boxes_to_draw=20,
                                          min_score_thresh=.5,
                                          agnostic_mode=False,
                                          line_thickness=4)
Improve this answer

6 Comments

Thanks for answering. I tried your hint but unfortunately results are same even with these attributes.
The question is what are the probabilities of the detected classes?
@Dat Tran No, I just want to return list or dictionary of objects that were detected, but with code I posted above it gives me dictioanary with 300 values, however there only few detected obejcts on image
I don't understand what you mean... Sure it only gives you a few detected objects because not all the values in the dictionary are really displayed. This depends on the probability threshold.
@Dat Tran my goal is to print to console only that number of objects or list of objects that is exactly found at image. I cant understand why num_detections returns 300 value or [category_index.get(i) for i in classes[0]] returns list with 300 values in it. Is it possible somehow to return list with number of objects that were found on image exactly?
|
1

Try to set the min_score_thresh to 0. Then you will probably see 300 detections.

Improve this answer

Comments

0

open visualization_utils.py and add--> print(class_name) after 

else:        
 class_name = 'N/A'
      display_str = '{}: {}%'.format(
          class_name,
          int(100*scores[i]))

this will print the detected objects

Improve this answer

Comments

0

adding print(class_name) after

else:        
 class_name = 'N/A'
      display_str = '{}: {}%'.format(
          class_name,
          int(100*scores[i]))

in visualization_utils.py file prints the detected object. I wonder where to add print command to print timestamps as well as percentage of accuracy in output.

Improve this answer

Comments

0 
// this will load the labels and categories along with category index

label_map = label_map_util.load_labelmap(PATH_TO_LABELS)

categories = label_map_util.convert_label_map_to_categories(label_map, max_num_classes=NUM_CLASSES, use_display_name=True)

category_index = label_map_util.create_category_index(categories)


//to print the identified object do the following :

print category instead of category index. The index holds the numeric value and the category contains the name of the objects. Once identified with the mentioned threshold the 

min_score_thresh = 0.5

print([category.get(1)] for i in classes[0] if scores[0, i] > min_score_thresh)

this will print the identified category.

Improve this answer

Comments

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.