Page 24 - Ai Robogenius

STEP - 2 Image Processing: The computer breaks the image into tiny dots called pixels. Then
it uses algorithms (smart instructions) to detect patterns and features in the image.

STEP - 3 Object Recognition: The system tries to identify what’s in the image—like a face,
object, shape or even text—by comparing it with what it has learned before (pre
trained data).
STEP - 4 Decision-making: Based on what it sees, the computer takes action. This could mean
moving a robot, raising an alarm, stopping a machine or making an independent
decision. For example,

At an international airport, a tech company is developing a facial recognition system
to enhance passenger security and speed up the check-in process. The system is
built in three main stages.

Training Analysis Interpretation

System analysis new System analysis new System interpretation
images, identifying key images, it identifies analysis data and makes
features and patterns key patterns a decision

In the first stage, the developers upload a large number of face images into the system.
These images include people of different ages, backgrounds and expressions. This helps
the system learn the key features of a human face, such as eye position, nose shape and
facial structure. This stage is known as training, where the system learns from a large
dataset to recognise different types of faces.
In the second stage, the system begins to analyse the new images it receives. It studies
important patterns and compares them with what it learned during training. It looks
for specific details that make each face unique, such as the distance between eyes or
the shape of the jawline. This stage is called analysis, where the system identifies useful
features from the input data.

In the final stage, the system uses its analysis to decide who the person is. When a new
traveller stands in front of the camera, the system compares the live image with its
memory and identifies the person. This stage is called interpretation, where the system
makes a decision based on the patterns it has recognised.

AI RoboGenius - VI
22