Vision System in Robots

Written by: Raunekk • Edited by: Lamar Stonecypher
Published Oct 31, 2009 • Related Guides: Electronic Devices

This article will explain as to how a robot using a vision system enables an object to be perceived and identified. The system uses various electronic devices to visualize a single object. Read more inside to find out how exactly this happens.

introduction

The vision system enables a robot to see and identify various objects. The system provides the robot with an ability to collect the light reflected from the objects, process an image out of it and then perform the desired action. The vision system helps with these functions by using various electronic devices and hardware systems.

How Does the System Works?

A vision system in a robot identifies an object by forming an electronic image using a bunch of pixels already stored in the memory of the robot’s controlling unit. Each pixel has a binary number allotted to it. Each of these binary numbers represents a particular wavelength and intensity in the light spectrum. An electronic image is formed in the controlling unit of the robot by assembling various binary numbers according to the amount of light.

Types of Vision Systems

A robot’s vision system is classified into three main types on the basis of the color of the objects. They are :

Binary image, which consist of black and white images
Gray colored images
Colored images with the base of red, green or blue.

An electronic image is formed with the help of pixels classified into these three categories. If an image is not been able to put in any of these categories, then the category that is extremely near to the image is selected.

Parts of the Process

A vision system will consists of a small camera, a monitoring system (a computer) and the necessary hardware and software. The whole process of identifying the image is classified into three main parts:

Image Processing
Thresholding
Connectivity paths

Image processing is a process by which an image is formed for analysis and use at a later stage. It uses various techniques such as image analysis and histogram of images to identify, simplify, modify or enhance an image.

Threshold is a process in which each image is classified into various categories and then compared with the pixels stored in the database. The pixels once compared are aligned to different levels to form an image.

The connectivity path is a process by which a particular pixel is connected to a neighboring pixel if it falls in the same color and texture region.

It is the combination of all these three processes that a final electronic image is conceived and the required action is taken after analysis.