Augmenting images with computer vision techniques
CNNs and computer vision are inseparable in deep learning. Before we dig deeper into the applications of deep learning for computer vision, we will introduce basic vision techniques that you can apply in your deep learning pipeline to make your model more robust. Augmentation can be used training to increase the number of distinct examples and make your model more robust for slight variations. Moreover, it can be used testing—Test Time Augmentation (TTA). Not every augmentation is suitable for every problem. For example, flipping a traffic sign with an arrow to the left has a different meaning than the original. We will be implementing our augmentation function with OpenCV.
How to do it...
- Let's first load all the necessary libraries:
import numpy as np import cv2 import matplotlib.pyplot as plt import glob
- Next, we load some sample images that we will use and plot them:
DATA_DIR = 'Data/augmentation/' images = glob.glob(DATA_DIR + '*') plt...