import numpy as np import cv2 import sys import matplotlib.pyplot as plt # Source: http://www.mattmahoney.net/ocr/ image1_path = "/work/NASDAQ_Example.jpg" img1 = cv2.cvtColor(cv2.imread(image1_path), cv2.COLOR_BGR2RGB)

fig, ax = plt.subplots(1, 2, figsize=(20, 10)) ax[0].imshow(img1) timg1 = (img1 < thresh).astype(np.uint8)*255 ax[1].imshow(timg1)

img2 = np.zeros((480, 640, 3), dtype=np.uint8) img2[100:200,100:300] = 255 img2[180:380,280:480] = 255 img2[240:320,340:420] = 0 fig, ax = plt.subplots(1, 3, figsize=(40, 10)) ax[0].imshow(img2) kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE,(kernel_width, kernel_width)) img2_dilate = cv2.dilate(img2, kernel) img2_erode = cv2.erode(img2, kernel) ax[1].imshow(img2_dilate) ax[2].imshow(img2_erode)

# Downsize image img1_noise = img1[::2, ::2, :].copy() shape = img1_noise.shape # Add noise for _ in range(200): i, j = np.random.randint(shape[0]-2), np.random.randint(shape[1]-2) img1_noise[i:i+1, j:j+2] = 0 for _ in range(200): i, j = np.random.randint(shape[0]-2), np.random.randint(shape[1]-2) img1_noise[i:i+1, j:j+2] = 255 fig, ax = plt.subplots(1, 2, figsize=(20, 10)) ax[0].imshow(img1_noise) timg_noise1 = (img1_noise < thresh).astype(np.uint8)*255 ax[1].imshow(timg_noise1) # Morphological operators kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE,(kernel_width2, kernel_width2)) timg_noise1_dilate = cv2.dilate(timg_noise1, kernel) timg_noise1_erode = cv2.erode(timg_noise1, kernel) fig, ax = plt.subplots(1, 2, figsize=(20, 10)) ax[0].imshow(timg_noise1_dilate) ax[1].imshow(timg_noise1_erode)

# Image source: https://upload.wikimedia.org/wikipedia/commons/thumb/d/d6/Purple_osteospermum.JPG/1599px-Purple_osteospermum.JPG image1_path = "/work/osteospermum.jpeg" # Image source: https://www.cs.rochester.edu/~brown/449/assts/fruit_tray.gif image2_path = "/work/fruit_tray.jpeg" img3 = cv2.cvtColor(cv2.imread(image1_path), cv2.COLOR_BGR2RGB) img4 = cv2.cvtColor(cv2.imread(image2_path), cv2.COLOR_BGR2RGB) fig, ax = plt.subplots(1, 2, figsize=(20, 10)) ax[0].imshow(img3) ax[1].imshow(img4) blur_kernel = cv2.getStructuringElement(cv2.MORPH_ELLIPSE,(blur_radius, blur_radius)) blur_kernel = blur_kernel / np.sum(np.abs(blur_kernel)) fig, ax = plt.subplots(1, 2, figsize=(20, 10)) ax[0].imshow(cv2.filter2D(img3, -1, blur_kernel)) ax[1].imshow(cv2.filter2D(img4, -1, blur_kernel))