What is Data Processing?

In this article, we will learn Data Processing,This free Tensorflow tutorial for complete beginners will help you learn Tensorflow from scratch.

How to use Data Processing in Tensorflow?

The given dataset is readily available in the TensorFlow datasets. Just download it from there using the tfds.load() function.

Data Processing

Downloading dataset

Visualizing Data

First, we will normalize the pixel values in the range of [0,1]. For this, we will divide each pixel value by 255. The dataset included from the TensorFlow is already divided into train and test split.

Python3

def normalize(input_image, input_mask):
   
    # Normalize the pixel range values between [0:1]
    img = tf.cast(input_image, dtype=tf.float32) / 255.0
    input_mask -= 1
    return img, input_mask
 
@tf.function
def load_train_ds(dataset):
    img = tf.image.resize(dataset['image'], 
                          size=(width, height))
    mask = tf.image.resize(dataset['segmentation_mask'],
                           size=(width, height))
 
    if tf.random.uniform(()) > 0.5:
        img = tf.image.flip_left_right(img)
        mask = tf.image.flip_left_right(mask)
 
    img, mask = normalize(img, mask)
    return img, mask
 
@tf.function
def load_test_ds(dataset):
    img = tf.image.resize(dataset['image'], 
                          size=(width, height))
    mask = tf.image.resize(dataset['segmentation_mask'], 
                           size=(width, height))
 
    img, mask = normalize(img, mask)
    return img, mask

Now we will set some constant values such as Buffer size, input height and width, etc.

Python3

TRAIN_LENGTH = info.splits['train'].num_examples
 
# Batch size is the number of examples used in one training example.
# It is mostly a power of 2
BATCH_SIZE = 64
BUFFER_SIZE = 1000
STEPS_PER_EPOCH = TRAIN_LENGTH // BATCH_SIZE
 
# For VGG16 this is the input size
width, height = 224, 224

Now, let’s load the train and test data into different variables and perform the data augmentation after batching is done.

Python3

train = dataset['train'].map(
    load_train_ds, num_parallel_calls=tf.data.AUTOTUNE)
test = dataset['test'].map(load_test_ds)
 
train_ds = train.cache().shuffle(BUFFER_SIZE).batch(BATCH_SIZE).repeat()
train_ds = train_ds.prefetch(buffer_size=tf.data.AUTOTUNE)
test_ds = test.batch(BATCH_SIZE)

Image Segmentation Using TensorFlow

Image segmentation refers to the task of annotating a single class to different groups of pixels. While the input is an image, the output is a mask that draws the region of the shape in that image. Image segmentation has wide applications in domains such as medical image analysis, self-driving cars, satellite image analysis, etc. There are different types of image segmentation techniques like semantic segmentation, instance segmentation, etc. To summarize the key goal of image segmentation is to recognize and understand what’s in an image at the pixel level.

For the image segmentation task, we will use “The Oxford-IIIT Pet Dataset” which is free to use dataset. They have 37 category pet dataset with roughly 200 images for each class. The images have large variations in scale, pose and lighting. All images have an associated ground truth annotation of breed, head ROI, and pixel-level trimap segmentation. Each pixel is classified into one of the three categories: