Computer Vision in Deep Learning

Introduction

In this blog we will explore idea of computer vision in deep learning before starting this blog you must understand the basic concepts of deep learning https://ainewgeneration.com/introduction-to-deep-learning/ before going in deep about computer vision you must know the basic concepts of computer vision and why it is most important nowadays. basically, computer vision is a field of AI that trained the computer to the visible whole world and understands what is happening in Images and Videos. So let’s get started in detail.

Table of Contents

1. What is Computer Vision ?
2. Computer Vision Applications
   • Object Classification
   • Object Detection
   • Object Landmark Detection
   • Object Segmentation
   • Object Recognition
   • Face Recognition

What is Computer Vision

Computer Vision simply means enabling the computer to identify and process objects, images, and videos in the same way a human does.

In the above image by looking as we humans are capable of understanding and describing the sense encapsulated in the image. when we look at the above image we will see the woman is crossing the road and also we see few cars on the road and trees. but when computers see the above image for the computer it is an array of pixels means these are numeric values that represent the shades of Red, Green & Blue (RGB).

Look at the image below what we see and how computers sees the same image in pixel numeric form.

Computer Vision Applications

• Object Classification
• Object Detection
• Object Landmark Detection
• Image Segmentation
• Object Recognition
• Face Recognition

Image Classification :

Object classification works only on a single object in an image. In the below image we have only one object in an image and simply we can classify that this image is of cat and in another side below image we perform classification and localization means we are classifying the image as what object is in it as well as we are detecting the bounding box where the particular object located in the image this is known as localization.

Object Detection :

Object Detection simply means when you have multiple objects in an image and you want to detect each and every object in an image then you simply go for object Detection Algorithm like YOLO, SSD, RCNN, etc we will discuss each algorithm in much detail in the coming blogs. YOLO object detection you can read the concept behind object detection and performing object detection using YOLO https://ainewgeneration.com/yolo-object-detection/.

Image Segmentation :

Image Segmentation means we want to know what pixel to the object in the image the different pixel belongs to cat and what are the different pixel belongs to dog object we will go much in details in coming blog about Image Segmentation.

Landmark Detection :

Landmark detection simply means finding key points for the objects in brief we will know about Landmark Detection in coming blogs.

Object Recognition

Object Recognition simply means what objects are in image/video and where they are. As you can see below image you are recognizing the object and where it is present in the image just drawing the bounding box and class to which it belongs to this is a brief of Object Recognition.

Face Recognition :

There are two things one is face detection and the other is face recognition. face detection simply means finding faces in an image whereas face recolonization simply means recognizing the faces of who is there in the image.

End Notes

I hope this blog gives you a clear basic idea about what is computer vision in Deep learning and what are their applications are. In the next blog, we will go into more detail regarding every application of deep learning we have discus in this blog.