Detect Objects Using Your Webcam

This demo will take you through the steps of running an “out-of-the-box” detection model to detect objects in the video stream extracted from your camera.

Create the data directory

The snippet shown below will create the data directory where all our data will be stored. The code will create a directory structure as shown bellow:

└── models

where the models folder will will contain the downloaded models.

import os

DATA_DIR = os.path.join(os.getcwd(), 'data')
MODELS_DIR = os.path.join(DATA_DIR, 'models')
for dir in [DATA_DIR, MODELS_DIR]:
    if not os.path.exists(dir):

Download the model

The code snippet shown below is used to download the object detection model checkpoint file, as well as the labels file (.pbtxt) which contains a list of strings used to add the correct label to each detection (e.g. person).

The particular detection algorithm we will use is the SSD ResNet101 V1 FPN 640x640. More models can be found in the TensorFlow 2 Detection Model Zoo. To use a different model you will need the URL name of the specific model. This can be done as follows:

  1. Right click on the Model name of the model you would like to use;

  2. Click on Copy link address to copy the download link of the model;

  3. Paste the link in a text editor of your choice. You should observe a link similar to;

  4. Copy the XXXXXXXXX part of the link and use it to replace the value of the MODEL_NAME variable in the code shown below;

  5. Copy the YYYYYYYY part of the link and use it to replace the value of the MODEL_DATE variable in the code shown below.

For example, the download link for the model used below is:

import tarfile
import urllib.request

# Download and extract model
MODEL_DATE = '20200711'
MODEL_NAME = 'ssd_resnet101_v1_fpn_640x640_coco17_tpu-8'
PATH_TO_CKPT = os.path.join(MODELS_DIR, os.path.join(MODEL_NAME, 'checkpoint/'))
PATH_TO_CFG = os.path.join(MODELS_DIR, os.path.join(MODEL_NAME, 'pipeline.config'))
if not os.path.exists(PATH_TO_CKPT):
    print('Downloading model. This may take a while... ', end='')
    urllib.request.urlretrieve(MODEL_DOWNLOAD_LINK, PATH_TO_MODEL_TAR)
    tar_file =

# Download labels file
LABEL_FILENAME = 'mscoco_label_map.pbtxt'
if not os.path.exists(PATH_TO_LABELS):
    print('Downloading label file... ', end='')

Load the model

Next we load the downloaded model

os.environ['TF_CPP_MIN_LOG_LEVEL'] = '2'    # Suppress TensorFlow logging
import tensorflow as tf
from object_detection.utils import label_map_util
from object_detection.utils import config_util
from object_detection.utils import visualization_utils as viz_utils
from import model_builder

tf.get_logger().setLevel('ERROR')           # Suppress TensorFlow logging (2)

# Enable GPU dynamic memory allocation
gpus = tf.config.experimental.list_physical_devices('GPU')
for gpu in gpus:
    tf.config.experimental.set_memory_growth(gpu, True)

# Load pipeline config and build a detection model
configs = config_util.get_configs_from_pipeline_file(PATH_TO_CFG)
model_config = configs['model']
detection_model =, is_training=False)

# Restore checkpoint
ckpt = tf.compat.v2.train.Checkpoint(model=detection_model)
ckpt.restore(os.path.join(PATH_TO_CKPT, 'ckpt-0')).expect_partial()

def detect_fn(image):
    """Detect objects in image."""

    image, shapes = detection_model.preprocess(image)
    prediction_dict = detection_model.predict(image, shapes)
    detections = detection_model.postprocess(prediction_dict, shapes)

    return detections, prediction_dict, tf.reshape(shapes, [-1])

Load label map data (for plotting)

Label maps correspond index numbers to category names, so that when our convolution network predicts 5, we know that this corresponds to airplane. Here we use internal utility functions, but anything that returns a dictionary mapping integers to appropriate string labels would be fine.

category_index = label_map_util.create_category_index_from_labelmap(PATH_TO_LABELS,

Define the video stream

We will use OpenCV to capture the video stream generated by our webcam. For more information you can refer to the OpenCV-Python Tutorials

import cv2

cap = cv2.VideoCapture(0)

Putting everything together

The code shown below loads an image, runs it through the detection model and visualizes the detection results, including the keypoints.

Note that this will take a long time (several minutes) the first time you run this code due to tf.function’s trace-compilation — on subsequent runs (e.g. on new images), things will be faster.

Here are some simple things to try out if you are curious:

  • Modify some of the input images and see if detection still works. Some simple things to try out here (just uncomment the relevant portions of code) include flipping the image horizontally, or converting to grayscale (note that we still expect the input image to have 3 channels).

  • Print out detections[‘detection_boxes’] and try to match the box locations to the boxes in the image. Notice that coordinates are given in normalized form (i.e., in the interval [0, 1]).

  • Set min_score_thresh to other values (between 0 and 1) to allow more detections in or to filter out more detections.

import numpy as np

while True:
    # Read frame from camera
    ret, image_np =

    # Expand dimensions since the model expects images to have shape: [1, None, None, 3]
    image_np_expanded = np.expand_dims(image_np, axis=0)

    # Things to try:
    # Flip horizontally
    # image_np = np.fliplr(image_np).copy()

    # Convert image to grayscale
    # image_np = np.tile(
    #     np.mean(image_np, 2, keepdims=True), (1, 1, 3)).astype(np.uint8)

    input_tensor = tf.convert_to_tensor(np.expand_dims(image_np, 0), dtype=tf.float32)
    detections, predictions_dict, shapes = detect_fn(input_tensor)

    label_id_offset = 1
    image_np_with_detections = image_np.copy()

          (detections['detection_classes'][0].numpy() + label_id_offset).astype(int),

    # Display output
    cv2.imshow('object detection', cv2.resize(image_np_with_detections, (800, 600)))

    if cv2.waitKey(25) & 0xFF == ord('q'):


Total running time of the script: ( 0 minutes 0.000 seconds)

Gallery generated by Sphinx-Gallery