International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 09 Issue: 05 | May 2022
p-ISSN: 2395-0072
www.irjet.net
ASSISTANCE SYSTEM FOR DRIVERS USING IOT Yogita Misal1, Atharva Agate2, Radhesham Joshi3,Pranav Bhosale4 1-3
Dept. of Information Technology, PES Modern College of Engineering Pune, Maharashtra, India
---------------------------------------------------------------------***---------------------------------------------------------------------------Abstract – This project entitled as “Assistance System for drivers using IOT” consisting of three major sections 1. Object Detection 2. Lane detection 3. IOT The evolution of Artificial Intelligence has served as the catalyst in the field of technology. We can now develop things which was once just imagination. This paper proposes a working model of assistance system which is capable of assisting drivers with a warning about possible obstacles and objects in front and the lane that the driver should follow. No matter how hard we try to create awareness regarding traffic rules and safety that has to be followed while driving, accidents are still occurring and aren’t showing a sign to stop. Though human errors can never be eliminated, but accidents can definitely be stopped. And in this case technology has surely come to our rescue. A camera module is mounted over the top of the car along with Raspberry Pi sends the images from real world to the Convolutional Neural Network which then detects and tracks the objects in-front and display them along with the distance between them.
INTRODUCTION we have created an assistive system for human drivers that tracks all the objects and lane in front of the car on the road by applying Computer Vision Techniques & convolutional neural network architectures on the frames generated through the live footage captured by the recording camera .Classifying images is straightforward, but the differences between object localization and object detection can be confusing, especially when all three are equally referred to as object recognition. Objects in images can be recognized by humans. Human vision is fast, accurate, and capable of identifying multiple objects and detecting obstacles with little conscious effort. We are now able to train computers with high precision to detect and classify multiple objects within an image with the availability of large data sets, faster GPUs, and better algorithms. We need to understand terms such as object detection, object localization, loss function for object detection and localization, and finally explore an object detection algorithm known as “You only look once” (YOLO). As well as assigning a class label to an image, image classification also involves drawing a bounding box around one or more objects in an image. A more challenging part of object detection is combining these two tasks to draw a bounding box around each object of interest in the image and assign it a class label. Object recognition is the result of combining all of these problems. An object recognition algorithm identifies objects in digital photographs by combining several tasks. R-CNNs, or region-based convolutional neural networks, are a family of techniques for recognizing objects and localizing them. A technique called You Only Look Once, or YOLO, is the second in a series of techniques for object recognition that uses speed and real-time processing. 1.1 OBJECT DETECTION Object detection is a critical vision task, but challenging at the same time. Image search, automatic annotation, scene interpretation, and object tracking are just a few applications that use it. Computer vision has focused a lot on tracking moving objects in video image sequences. In addition to smart video surveillance (Arun Hampapur 2005), artificial intelligence, guidance for military vehicles, safety detection, and robot navigation, and medical and biological applications, it has already been applied in many fields of computer vision. In recent years, there has been a number of successful single object tracking systems, however object detection becomes difficult when there are more than one object in the scene and when objects are occluded, their vision is obscured, which further complicates detection.
© 2022, IRJET
|
Impact Factor value: 7.529
|
ISO 9001:2008 Certified Journal
|
Page 574