Implementation Of Walking Aid For Visually Impaired People On Accelerated System
This project is for visual Impaired people to give them an aid which they would be able to use for mobilization. We want to use technology to make a difference in the lives of visual Impaired community. Our aim is to build a system which will ease these people?s movement from one location to another
2025-06-28 16:33:04 - Adil Khan
Implementation Of Walking Aid For Visually Impaired People On Accelerated System
Project Area of Specialization Electrical/Electronic EngineeringProject SummaryThis project is for visual Impaired people to give them an aid which they would be able to use for mobilization. We want to use technology to make a difference in the lives of visual Impaired community. Our aim is to build a system which will ease these people’s movement from one location to another. These people face issue when they have to move from location to another either they are dependent on a person or a walking cane which does not alert them when a person or an object is in front of them and where they are right or in an emergency what they would do.
So, our walking stick would detect obstacles and people. We will use this data to alert either the object or the user of this aid is at a safe distance from the object. If the user is in an emergency, he would be able to send his location to anyone..
Project Objectives- Facial and Object recognition with real time enhancement of
- Data Set Hurdle detection and Avoidance
- Location Sharing
We are using NVIDIA Jetson Nano Developer kit as our accelerator system. Which will contain Docker Container which will conatin the dataset and trained model SSD (Single Shot Detector) MobileNetV2 which we will be used to detect objects and do facial detection. Video would be recorded through the Camera attached to the accelerator system. Code of the SSD (Single Shot Detector) MobileNetV2 is written in Python Programming Language and Deep learning frame work which has been used is PyTorch.To optimized the neural network layers NVIDIA TensorRT is used for faster Inference during the run time. NVIDIA TensorRT is built on the NVIDIA CUDA for parallel computing. We will be using USB Wireless adapter to connect this accelerator system to Internet.In case of emergency user would be able to push a button to send his current location to his trusted person for help. we have used ultrasonic Sensor for hurdle detection if a hurdle would be detected then the buzzer and the Speaker would be utilized to inform the user about hurdle and to avoid him.
Benefits of the ProjectThis project would be really beneficial for Industry because this project is using Computer Vision Algorithm SSD (Single Shot Detector) MobileNetV2 which can be used to detect Objects and faces.
Currently Laser beam technology is being used to detect objects in a cane or a smart glasses which like you can see in this video which is sending captured picture from Camera inside the smart glasses and it’s using Microsoft Azure Computer Vision Cognitive Service to tell what objects are in the Image it goes to Cloud to process this Image which is adding cost and also very latent.
So, our Solution is to use accelerated system in the edge rather than going to cloud which would be very latent. Our project has very vast applications for Industry like our computer vision algorithm after training with appropriate data set could be used to do surveillance and monitoring of the traffic.
It could also be used in Self-Driving Cars to detect their surrounding and take action according to the circumstance.It’s application is also in a manufacturing industry to maintain their quality we can use this project computer vision algorithm to detect defective products and separate them.
Technical Details of Final DeliverableOur final Project would be to develop an Portale device which could be attached with an Walking Aid and help the blind person to move from one location to another.
We are using Nvidia Jetson Board which has GPU in it which has 128 Cores. Our algorithm would be depolyed in it and running in it and using CUDA software for parallel computing.
This algorithm would do object detection and facial recognition.In case of the emergency user can push a button location of our user would be shared.
Final Deliverable of the Project Hardware SystemCore Industry HealthOther Industries IT Core Technology Artificial Intelligence(AI)Other Technologies Cloud InfrastructureSustainable Development Goals Good Health and Well-Being for People, Industry, Innovation and InfrastructureRequired Resources| Item Name | Type | No. of Units | Per Unit Cost (in Rs) | Total (in Rs) |
|---|---|---|---|---|
| Total in (Rs) | 60350 | |||
| Jetson Nano Board | Equipment | 1 | 24500 | 24500 |
| Ultra Sonic Sensor | Equipment | 2 | 200 | 400 |
| Raspberry Pi Camera Module V2 | Equipment | 2 | 4500 | 9000 |
| TL-WN 725N | Equipment | 1 | 1300 | 1300 |
| Imported Brand Rollator Wheel Walker For Seniors and Patients | Equipment | 1 | 13900 | 13900 |
| Services and Consultation | Miscellaneous | 1 | 10000 | 10000 |
| USB 5.5 mm Female Jack Adapter | Equipment | 1 | 250 | 250 |
| 64 GB SAMSUNG MEMORY | Equipment | 1 | 1000 | 1000 |