Apache Spark Deep Learning Cookbook

How to do it...

The steps are as follows:

1. Define the crop() function to crop images to obtain only the significant part, as shown in the following lines of code:

 #function for cropping images to obtain only the significant part def crop(img):      a=28*np.ones(len(img))       b=np.where((img== a).all(axis=1))       img=np.delete(img,(b),0)       plt.imshow(img)      img=img.transpose()      d=28*np.ones(len(img[0]))      e=np.where((img== d).all(axis=1))      img=np.delete(img,e,0)       img=img.transpose()      print(img.shape)       super_threshold_indices = img < 29       img[super_threshold_indices] = 0      plt.imshow (img)      return img[0:150, 0:128]

2. Use the following lines of code to loop through every image in the folder and crop it using the preceding defined function:

#cropping all the images ...