Robot Automation through Image Processing

The main aim of robot automation through image processing is that, it does not require any sensor in order to operate a robot. So we are not bounded to operate robot from specified distances. Moreover robot automation is a time consuming and repeating process in which we have to provide command one

Project Title

Robot Automation through Image Processing

Project Area of Specialization

Robotics

Project Summary

The main aim of robot automation through image processing is that, it does not require any sensor in order to operate a robot. So we are not bounded to operate robot from specified distances. Moreover robot automation is a time consuming and repeating process in which we have to provide command one by one in order to operate robot manually.So according to this idea robot is operated automatically through machine vision. An algorithm is designed for processing it, in which instructions are not provided again and again for motion detection. Rather it takes input from person standing in front of webcam. Movement of a person is continuously monitored and through this live feed movements are detected by using algorithms of connected component analysis. This process automatically obtains positions and instructions are given accordingly to operate robot. Robot is integrated with software that will automatically provide instructions to generate pulse width modulated waveforms to control speed of servo motors and in turn move robotic arm.

Project Objectives

Many organizations are switching to robot automation instead of manual labor work to streamline their processes with future technology that is advancing day by day in the field of robotics. The research is still in progress to achieve high versatility, high value and high robustness in the field of automation.Moreover a human operator can control multiple robots with support of automation. So overall human robot performance is considered as higher that reduces workload. Moreover industrial robots are becoming essential part in order to compete global markets that can easily be configured again according to the needs.Robotic systems are used for a variety of purposes includes manufacturing, surgery, automation, surveillance.To operate robot we have to reprogram them according to task we desired them to perform. What if we needed them for multiple tasks which changes uncertainly? We are not sure how to program it manually. In such case we need a way to program it in such a way that it changes according to user’s desire. For this computer vision is best technique to adapt as once we designed the algorithm to detect movements we do not need extra time to program it for operations and not limited to predefined tasks.

Project Implementation Method

The overall project is consisted of two modules. One deal with software and another is software hardware integration .The first module of design in which live feed is continuously captured from webcam. Then it is processed by using Python software that is specifically designed for image processing. Python first capture live stream, then background is removed  by using background subtraction methods. A threshold is selected to remove the background from entire image. Then this image is separated from entire frame.  connected component analysis technique is then applied to seperate the label and the labels across hand are seperated to find the edges and joints as person moves. In order to detect edges we need a threshold to find the difference between edges and entire background. Outer borders and hole borders have one to one correspondence so same pixel value is assigned to both. So features are extracted from image without reconstructing.After detecting edge, to find joints and other movements, a straight line is required that can define the proper path for robot to automate.For this the mean of borders drawn is taken and points are obtained. Angle is calculated at joints.The second module of design in which robotic arm is controlled by following instructions. By serial communication instructions are given to the controller of robot that uses uno arduino. Four servo motors are interfaced by this controller. Their speed is controlled by changing pulse width of waveforms provided.­

Benefits of the Project

It will automate robot in such a way that can be modified with user's input without programming it again and again. so it can prove helpful for various application.For example it can be used in medical field as a surgeon want to conduct any surgery through overseas by giving instructions to robot. It can be utilized in the field of education to mimic for deaf students.

Technical Details of Final Deliverable

• Movement detection through image processing in python software
• To operate robotic arm with four servo motors
• Software hardware intyegration by serial communication

Final Deliverable of the Project

HW/SW integrated system

Type of Industry

Education , Medical , Others

Technologies

Robotics

Sustainable Development Goals

Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com