OpenCV 3 Computer Vision with Python Cookbook

How to do it...

1. Import the modules:

import cv2import matplotlib.pyplot as plt

2. Load the test image:

image = cv2.imread('../data/people.jpg')

3. Create the HOG feature computer and detector:

hog = cv2.HOGDescriptor()hog.setSVMDetector(cv2.HOGDescriptor_getDefaultPeopleDetector())

4. Detect the people in the image:

locations, weights = hog.detectMultiScale(image)

5. Draw the detected people bounding boxes:

dbg_image = image.copy()for loc in locations:    cv2.rectangle(dbg_image, (loc[0], loc[1]),                   (loc[0]+loc[2], loc[1]+loc[3]), (0, 255, 0), 2)

6. Visualize the results:

plt.figure(figsize=(12,6))plt.subplot(121)plt.title('original')plt.axis('off')plt.imshow(image[:,:,[2,1,0]])plt.subplot(122)plt.title('detections')plt.axis('off')plt.imshow(dbg_image[:,:,[2,1,0]]) ...