Smart Vision Goggles for Blinds

Smart Vision Goggles for Blinds

ABSTRACT

In this project, we are making a wearable device to help visually impaired people in a social interaction. It is one of the most demanding tasks for the people who are blind or have low vision to recognise in a various social interactions. To identify the people recognition through voice is the most commonly used method but in this project we are adding some more features like use of Amazon Rekognition that is used for identify classmates, friends and relatives by giving a message on earphone to give notification to the user and has a facility to tag image by giving some identity or complete information for new people. If the same people appear in front of the user then the complete information of a person is given as a message on an earphone. We place any book or newspaper in front of the camera, it will capture the image and process it and speech synthesis converts that text into speech. This way, user can also take advantage of this device when he/she wants to read any newspaper or book. This proposed device supports the blind in every way to recognise people, read everything and it also makes them independent.

OVERVIEW

Could you imagine the life of a blind person? It is very difficult for them to live alone and they are totally depending on their care-givers. So that’s why we just want to help those people by making “Smart Vision goggles for blind” project using Amazon Rekognition and Raspberry Pi which is the best way to help the bind or low vision people to recognise the things and become self- dependent. We did the deep research about this project on various resources like research databases, books and other internet resources. In this project, these smart goggles helps the people to walk alone, live their life independently by taking inputs through an Amazon rekognition service and providing response to the blind person through headphones. So, visually impaired people can be trained to visualize the various objects. Smart goggles can be worn in front of the eyes same as other normal goggles, but the difference is just that it is a smart computing gadget.

Our project has two scopes:-

1)      Image Detection and Recognition

2)      Smart Reader and IR Obstacle Detection

Block diagram of smart vision goggles for blind project:-

Image Detection and Recognition: – In this scope of our project, we are using Amazon Rekognition which can be used for identifying classmates, friends and relatives and information can be heard through an earphone to give notification to the user and has a facility to tag image by giving some identity or complete information for new people. In this part, camera will capture the image that is in front of them then image processing will convert an image into the digital form and perform some operations on it to get an enhanced image then detect and recognise the Image and then send that image to the speech synthesis that will covert that image into speech and provide that voice information to the user through headphones. (AWS)

Smart Reader and IR Obstacle Detection: – smart reader is a method which is helpful for the blind people to read newspapers, books and other things. In this, if we place anything like book magazine in front of the camera, the camera will capture the image process it and convert that printed text into the speech form and provide them that voice information to the user through headphones. This method helps the blind people to read street signs, restaurant menus, newspapers and food labels. Besides that in this project we are also using IR obstacle detection circuit that will detect the obstacle that came in front of the user and alert the user by making beep sound. (Amr sarawy)

Components required in this project are given below:-

      Raspberry Pi 3 B

      Camera Module

      IR Obstacle Detection Circuit

      Headphone

      MicroSD Card

      Sim Card/USB Modem

      Battery

Technologies used in this project are given below:-

      Amazon Rekognition

      Speech synthesis

      Image processing

      Python Coding

Objectives: –

1.)    Our main motive of this project is to help the visually impaired people, not accurately but to make their life little bit easier and become self- dependent.

2.)    In this project, the glasses we will be using would be able to take pictures via camera.

3.)    “Glasses can recognise images and determine each object in the images”.

4.)    It can determine the distance between the blind person and each object.

5.)    Conversion of captured image information into the voice will be provided to user through headphones that help the blind people to know who is in front of his/her.

6.)    It will also give notification to user if the object is very close or far away from him/her.

Advantages: –

1.)    Smart goggles are very easy to use and carry.

2.)    Smart glasses use various softwares, so that’s why we are using Amazon Rekognition to make the project cheaper than other similar projects.

3.)    Amazon rekognition helps blind people a lot to recognize their friends, relatives, vehicles, celebrities and other objects and we are also using an IR Obstacle detection circuit which will alert the user if there will be any kind of obstacle in his/her way heading towards.

4.)    Deep learning-based image analysis – Amazon Rekognition uses deep-learning technology to accurately analyze images, find and compare faces in images, and detect objects and scenes within your images.

5.)    It enable the user to read text books and news papers as well.

Project scope: –

•          This project has a great scope and huge demand for the visually impaired people in today’s world. As according to the world Health Organisation almost 124 million people in this world are suffering from blindness. So, that’s why we choose this project to help these kind of people and make them self-dependent. The ‘Smart Vision Goggles’ are advanced technology and programmed wearable device which can allow and assist blind or visually impaired people while walking alone in new environments by taking inputs through a camera and providing feedback to the person through headphones. In short words These ‘Smart Vision Goggles’ can help visually impaired to ‘see’ or ‘feel’ the world virtually by using camera, face recognition software and sensors.

Work Breakdown Structure (WBS):-

Project Methodology and Requirements:-

 

S. No.	Activity Description	Accountability	Role	Technology
1	Project Proposal Preparation and Presentation	All the Team members	Presenter	Project Libre
2	Project Plan	All the Team members	Planner	Project Libre
3	Material Procurement (H/W)	Kirandeep Kaur	Material Procurement and Inventory Regulator
4	S/ W arrangements	Amit Kumar	System Administrator
5	H/W and S/W installation	All the Team members	System administrator and programmer	Amazon Rekognition, Speech Synthesis and Python Coding
6	AWS Database Creation	Amit Kumar	Database Developer	RDS MySQL
7	IR sensor circuit and Text Reading	Kirandeep Kaur	System Engineer and Programmer	Infrared technology and TTS
8	Database Testing	Amit Kumar	Tester	Amazon DynamoDB and AWS Lambda
9	Camera Testing	Kirandeep Kaur	Tester	Raspbian
10	Monitoring (IR Sensor and Text Reader Scope)	Kirandeep Kaur	Project In-charge	–
11	Monitoring (Face Rekognition)	Amit Kumar	Project In-charge	–
12	Final Project Presentation	All the Team members	Presenter	–
13	Project Submission	All the Team members	–	–

 

 

The Explanation that how the project will be carried out: -We are two members in a group. So we will work together and complete the project on time. As we will do two activities in our project and from those activities I will do the text to speech conversion project and my partner will do the image to speech conversion using amazon recognition. As we assumed that we will complete our project in 10 weeks.

Required equipments and components list with their cost: – Below given list describes the required equipments and components along with their respective and overall cost (approx.): –

Hardware Components	Cost in NZD (Approx.)
      Raspberry Pi	100
      Camera Module	35
      IR Obstacle Detection Circuit	30
      Headphone	20
      MicroSD Card	30
      Sim Card/USB Modem	35
      Battery	30
      Miscellaneous	50
      Overall Cost Estimation	330

Software requirement list for this project is as given below:-

      Amazon Rekognition

      Speech synthesis

      Object recognition

      Image processing

Below given Risk matrix reveals the short note about the possible risks that may occur during the project progression:

 

Potential Risks and Mitigation techniques: –

Risk	Mitigation	Seriousness	Likelihood	After implementation change
Project purpose and need is not well-defined	Pre-project study and analysis , Project charter preparation, Scope defining	High	Medium
Strategic Risk	Strategy Management Plan	High	Medium
H/W or S/W Failure	Reserve backup	High	High
Team Communication Failure/Conflicts among Team members	Communication Management and job allocation	High	Medium
Budget Risk	Cost Planning and management, Monitoring and controlling Budget	High	Medium
Schedule Delay	Create Gantt Chart, define activities and sub-activities	High	High
Resource Risk	Resource Management Plan	High	Medium

Project Plan and Schedule: –

We have developed below given activity sheet and Gantt chart to meet the time schedule requirements and complete the project work on time:

Works Cited

• Amr sarawy, Y. E. (n.d.). Smart Glasses for blinds. Retrieved September 15, 2019, from www.academia.edu: https://www.academia.edu/33796786/Smart_glasses_for_blind_people
• AWS. (n.d.). What Is Amazon Rekognition? Retrieved September 15, 2019, from www.docs.aws.amazon.com: https://docs.aws.amazon.com/rekognition/latest/dg/what-is.html