Yolo Based object detection and distance estimation using Jetson nano

In recent years, autonomous vehicle industries have received great attention due to advancement in artificial intelligence field.Traditional vehicles can be improved with advances in artificial intelligence and its applications in various fields. Automotive industries are developing selfdriving vehi

Project Title

Yolo Based object detection and distance estimation using Jetson nano

Project Area of Specialization

Artificial Intelligence

Project Summary

In recent years, autonomous vehicle industries have received great attention due to advancement in artificial intelligence field.Traditional vehicles can be improved with advances in artificial intelligence and its applications in various fields. Automotive industries are developing selfdriving vehicles equipped with driver's safety technology. To cope with increasing demand of recognition technology for autonomous vehicles, we build robust recognition system to assist driverless vehicles to navigate safely without compromising safety of passengers and pedestrians. The proposed system detects objects and estimates how far, absolute distance, they are from the vehicle. The system is consisted of jetson Nano kit, equipped with essential peripherals, that add human-like intelligence to autonomous vehicle to take correct decisions at time of changing lane, deaccelerating, accelerating and applying the brakes. The system exploits Yolo model, a recognition algorithm, to recognize different objects in the image captured by the camera. The automotive industries are in race of designing most advanced driverless vehicles by applying sophisticated AI techniques and integrating advanced sensors.

Initially, LiDAR sensing technology is explored to be used in vehicles to detect objects. Although LiDAR is more accurate but is expensive. On the contrary, camera, very cheap vision sensor, captures sufficient information and enables the recognition algorithm

Project Objectives

The global autonomous vehicle market was valued at $76.13 billion in 2020, and is projected to reach $2,161.79 billion by 2030. Our objective is to build cheap objection detection and distance estimation system for autonomous vehicle. Autonomous vehicle’s distance estimation is an essential part of 3D scene recognition and orientation, allowing various autonomous devices to move in the natural environment. Autonomous vehicle can be equipped with LiDar, Radar which are expensive. So we use cameras for distance estimation powered with Computer vision algorithm. This project creates a system capturing a scene in front of a car that detected objects with their absolute distances by YOLO using Jetson nano

Project Implementation Method

YOLO (you only look once) based object detection and distance estimation system capture a scene in front of a car and the detected objects with their absolute distances. 1: Product development of project software and hardware to meet the international standards for commercial feature to be used in autonomous vehicles.

2:Monoclular Camera required to capture scene on road YOLO (You Only Look Once) based system, process it and detect the object with absolute distance estimation using Jetson nano.

3: Proper setup for autonomous vehicle companies and solution to build self-driving cars vision to manufacture in Pakistan

4: Targeting the Global Autonomous Vehicle Manufacturers to get this efficient and cheap AI-powered vision for autonomous vehicle

Benefits of the Project

Yolo Based object detector and distance estimation using Jetson nano system will provide benefits

Creating 3D Maps

It will enable self-driving vehicles to capture visual data in real time. The cameras attached with such vehicles can record live footage and allow computer vision to create 3D maps. Using these maps, autonomous vehicles can understand their surroundings better while spotting obstacles in their path and opt for alternate routes with 3D maps.

Computer Vision-Enabled Low-Light Mode

In order to process low light images and videos, self-driving vehicles use different algorithms than the ones used for daylight. The images captured in low light may be blurry and such data may not be accurate enough for these vehicles.

More Efficient

Cameras will do a better job traversing that environment than other sensors, reading all the vision queues meant for drivers and like road signs.

Less Expensive

This project is less expensive because it uses cameras for object detection and distance estimation.

This system can be attached to any autonomous vehicle like cars, drones many more

Technical Details of Final Deliverable

This project is made in steps:

1: Building the Deep Learning model using YOLO for object detection and classification present on roads. Then also achieving the distance estimation of that objects using cameras.

2: Setting up the jetson nano for project and downloading OS and all software modules that required for Deep Learning Model. Attaching the monocular camera with jetson nano so that it can capture the scenes in front of car

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Transportation

Other Industries

IT , Security

Core Technology

Artificial Intelligence(AI)

Other Technologies

Internet of Things (IoT), Clean Tech

Sustainable Development Goals

Industry, Innovation and Infrastructure, Life on Land

Required Resources

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