Artificial Vision System for Blind
Nowadays blind people use their conventional stick to identify the obstacles in front of them. They use the stick to touch and feel the obstacle, unevenness of pathway, staircase. These are difficult on a daily basis and hinders their mobility. They also have to seek help from bystanders when moving
2025-06-28 16:30:18 - Adil Khan
Artificial Vision System for Blind
Project Area of Specialization Mechanical EngineeringProject SummaryNowadays blind people use their conventional stick to identify the obstacles in front of them. They use the stick to touch and feel the obstacle, unevenness of pathway, staircase. These are difficult on a daily basis and hinders their mobility. They also have to seek help from bystanders when moving from place to place or while crossing the road.
Artificial Vision System for Blind will ensure safety for the blind while walking. This device will be carried by blind and when the device will sense any obstacle it should give the message through voice processor and accordingly, he can move. The entire design is built on the concept of an embedded system and the device will be helpful for blind people to go on an exact path without the help of others.
By using an intelligent device, visually impaired people can improve their travel speed and can reduce minor collision as well as double the safety. Visually disabled people use the stick for walking purposes as it provides such features with the help of an ultrasonic sensor which will be used for object detection, water detection will be done by using the moisture sensor, and atmospheric light detection by using an LDR sensor. It'll capture the images of detected obstacles, recognize the objects in front of the user, and provide the information with the help of speaker or headphones. It will also identify the currency notes when the user places a paper currency in front of the mounted camera. This type of system is based on the real-time embedded system because it is a modern technology that depends upon the hardware, software, and other parts.
Project Objectives- The main objective of this project is to provide safety for blind people from
obstacles, using an "Intelligent system" that can recognize/identify the objects coming in their way. - The clear path indication and environment recognition is also the aim of this system.
- The project must also perform:
- Object recognition using image processing.
- Currency recognition
- Image-to-Speech conversion (for images containing any textual information)
- Location detection using GPS/GSM modules.
There are two modes in the system namely WALK mode and CAM mode. In WALK mode, AVSB uses the information acquired with the help of ultrasonic sensors to determine the distance of the obstacle from the user and guide him towards the best possible path thereby avoiding obstacles on the way.
CAM mode, in addition to ultrasonic sensors, utilizes information from the camera for facilitating the blind man in recognizing various objects. In both these modes, the user is given information about the path and the object through voice commands. An ear speaker attached to the system makes these voice commands available for the user.
Algorithm:
WALK mode algorithm is simple and can be implemented in C language. As long as WALK mode is switched on; ultrasonic sensors keep on finding the obstacles and the user is being guided towards the available path. When CAM mode is switched on control is transferred to the Raspberry Pi for object recognition.
Obstacle detection using Ultrasonic Sensors:
When the blind person is walking, the sensors will detect any obstacle coming in their way. The output of this sensor goes to the microcontroller. The output of the controller is then given to the Raspberry Pi kit which gives the output to the user in the form of the audio signal as an object is sensed through the earphones/speakers.
Object Recognition using Image processing:
When the user enters someplace e.g. a room, he/she will press a specified button on the keypad mounted on the device. This will call a function in which an ANN algorithm is performed for the recognition of the image captured. The output in the form of an audio signal tells the name of the object to the user.
Image to Speech Output using Image processing
The blind person can place the image (in which some text is written) in front of the mounted webcam and press a specified button on the keypad. This will call a function in which image processing is performed to convert it into text. This text will be sent to Raspberry pi to convert into an audio signal. The output sent to the user via earphones. In this way, blind people will be able to hear the written content rather than read.
Currency Detection using image processing:
The user can place paper currency in front of the mounted webcam and press a specified button. This will call a function and python coding will be performed to detect the amount of that currency. The output will be given to the user via earphones.
Location detection using GPS/GSM:
GPS module is used to know the current location where the blind person is present, he can also hear the audio message regarding the direction that is to be followed by the blind person. GSM module is used by the blind person to contact mobile numbers stored in the microcontroller in case of any emergency, i.e. If the visually impaired person also wants some help during some emergency, a call or a message is sent to a set of mobile numbers stored in a microcontroller.
Benefits of the Project- The Artificial Vision system for Blind is integrated with multiple functions and this will aid Blind people in performing their daily chores easily, independently, and more efficiently in an affordable manner.
- One of the major factors will be the cost of this system. Since blind people are already paying for treatments or other nursing care, the price of this system would be in the range that is reasonable.
- Another feature of this product would be the durability. The users might not be able to charge the system. So, appropriate measures will be taken for it.
The Artificial Vision System for Blind is an Intelligent system that is based on the real-time embedded system because it is a modern technology that depends upon the hardware, software, and other parts. An embedded system is normally based upon the real-time operating system, various types of micro-controllers, programming language, arithmetic instruction set, and cost-conscious market, and with the help of GPS (Global Positioning System), we can easily achieve useful navigation in an outdoor application like identifying the location, position, and orientation of visually disabling person.
| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 40200 | |||
| Raspberry Pi 4 Model B | Equipment | 1 | 15000 | 15000 |
| Arduino Mega | Equipment | 1 | 1500 | 1500 |
| Ultrasonic Sensors | Equipment | 4 | 300 | 1200 |
| PiCam 8MP Module 2 | Equipment | 1 | 3000 | 3000 |
| GPS/GSM Module with antenna | Equipment | 1 | 4000 | 4000 |
| Power supply (10000mAh) | Equipment | 1 | 5000 | 5000 |
| D2A converter | Equipment | 1 | 1500 | 1500 |
| Accessories (Earphones, keypad, wires, etc) | Equipment | 1 | 2000 | 2000 |
| Thesis printing + overheads | Miscellaneous | 1 | 7000 | 7000 |