Image Classification Pipeline

A computer vision pipeline is a series of steps that most computer vision applications will go through. Many vision applications start off by acquiring images and data , then processing that data, performing some analysis and recognition steps, then finally performing an action. The general pipeline is pictured below:

Given sets of input images, computer vision techniques are used to process those images, identify images and characteristics of interest, and then use that data to recognize certain objects and interpret an image. In this section, we will be focusing on learning more about each individual step in this pipeline.

We will start by learning about how a single image is formed and digitally represented. Then we'll get some practicing on pre-processing techniques and selecting areas of interest in an image.

Image Classification Pipeline

An image classifier is an algorithm that takes in an image as input and outputs a label or “class” that identifies that image. For example, a traffic sign classifier will look at different of roads and be able to identify whether that road contains humans, cars, bikes and so on. Distinguishing and classifying each image based on its contents.

There are many types of classifiers, used to recognize specific objects or even behaviors — like whether a person is walking or running — but they all involve a similar series of steps...

1. First, a computer receives visual input from an imaging device like a camera. This is typically captured as an image or a sequence of images.
2. Each image is then sent through some pre-processing steps whose purpose is to standardize each image. Common pre-processing steps include resizing an image, or rotating it, to change its shape ortransforming the image from one color to another - like from color to grayscale. Only by standardizing each image, for example: making them the same size, can you then compare them and further analyze them in the same way.
3. Next, we extract features. Features are what help us define certain objects, and they are usually information about object shape or color. For example, some features that distinguish a car from a bicycle are that a car is usually a much larger shape and that it has 4 wheels instead of two. The shape and wheels would be distinguishing features for a car. And we’ll talk more about features later in this lesson.
4. And then, finally, these features are fed into a classification model! This step looks at any features from the previous step and predicts whether, say, this image is of a car or a pedestrian or a bike, and so on.

Image Pre-processing >>