Paso 1

Configuracion de Object Detection API: en este paso, se descarga el modelo para la detección de objetos, también se realizan algunas copias y eliminaciones de referencia con el objetivo de dejar todo el ambiente configurado.



In [0]:

    
!git clone https://github.com/tensorflow/models.git
!apt-get -qq install libprotobuf-java protobuf-compiler
!protoc ./models/research/object_detection/protos/string_int_label_map.proto --python_out=.
!cp -R models/research/object_detection/ object_detection/
!rm -rf models

Paso 2

Importaciones necesarias para ejecutar la demostración de Object Detection API



In [0]:

    
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

Paso 3

Configuración del modelo a utilizar, ruta al modelo pre-entrenado y elementos de configuración adicionales para la implementación de Object Detection API.



In [0]:

    
MODEL_NAME = 'faster_rcnn_inception_resnet_v2_atrous_coco_2018_01_28'
MODEL_FILE = MODEL_NAME + '.tar.gz'
DOWNLOAD_BASE = 'http://download.tensorflow.org/models/object_detection/'
PATH_TO_CKPT = MODEL_NAME + '/frozen_inference_graph.pb'
PATH_TO_LABELS = os.path.join('object_detection/data', 'mscoco_label_map.pbtxt')
NUM_CLASSES = 90

opener = urllib.request.URLopener()
opener.retrieve(DOWNLOAD_BASE + MODEL_FILE, MODEL_FILE)
tar_file = tarfile.open(MODEL_FILE)
for file in tar_file.getmembers():
  file_name = os.path.basename(file.name)
  if 'frozen_inference_graph.pb' in file_name:
    tar_file.extract(file, os.getcwd())
    
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)

Paso 4

Sección con las imágenes de demostración



In [ ]:

    
!mkdir images
# esta url-imagen debería ser reemplazada por ustedes. este es solo el ejemplo almacenado en una capeta personal
!wget https://storage.googleapis.com/demostration_images/image.jpg -O images/image_1.jpg

PATH_TO_TEST_IMAGES_DIR = 'images'
TEST_IMAGE_PATHS = [ os.path.join(PATH_TO_TEST_IMAGES_DIR, 'image_{}.jpg'.format(i)) for i in range(1, 2) ]
IMAGE_SIZE = (15, 11)

Paso 5

Pieza de implementación que representa la detección concreta, llamando a la sesión TF



In [0]:

    
with detection_graph.as_default():
  with tf.Session(graph=detection_graph) as sess:
    image_tensor = detection_graph.get_tensor_by_name('image_tensor:0')
    detection_boxes = detection_graph.get_tensor_by_name('detection_boxes:0')
    detection_scores = detection_graph.get_tensor_by_name('detection_scores:0')
    detection_classes = detection_graph.get_tensor_by_name('detection_classes:0')
    num_detections = detection_graph.get_tensor_by_name('num_detections:0')
    for image_path in TEST_IMAGE_PATHS:
      image = Image.open(image_path)
      image_np = load_image_into_numpy_array(image)
      image_np_expanded = np.expand_dims(image_np, axis=0)
      (boxes, scores, classes, num) = sess.run(
          [detection_boxes, detection_scores, detection_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=3)
      plt.figure(figsize=IMAGE_SIZE)
      plt.imshow(image_np)