Abdullah Ali

Academic

Career

Verbatim text

Computer Vision Based Detection of Zero to Low Visibility Conditions 
his project aims to develop a real-time fog and smog detection system to enhance road safety under 
low-visibility conditions. Fog and air pollution significantly reduce visibility, increasing the risk o 
ccidents on highways and in urban areas. The proposed system intelligently monitors environmental 
onditions and assists drivers by providing timely warnings during hazardous situations. 
he system is built using Jetson Nano and Raspberry Pi-5 as the main processing unit and integrates 
camera, radar sensor, and thermal sensor to ensure reliable performance in different weather and 
lighting conditions. It leverages the concepts of sensor fusion and sensor switching, where data from 
multiple sensors are combined to improve detection accuracy and reduce false readings. In dense fog 
r heavy smog, when the optical camera's performance is affected, the system automatically 
prioritizes radar or thermal sensing to maintain effective detection and obstacle awareness. 
By applying computer vision and deep learning techniques along with multi-sensor data processing, 
he system analyzes visibility levels and environmental risks in real time. The camera feed is processed 
using YOLO v26, a deep learning object detection model, to detect vehicles, pedestrians, and other 
bstacles efficiently. When unsafe conditions are detected, the system generates immediate alerts to 
ssist the driver in taking precautionary actions. 
his intelligent multi-sensor approach contributes to the advancement of driver assistance 
echnologies and smarter transportation safety systems. By continuously adapting to environmental 
onditions and providing timely warnings, it significantly enhances road safety and supports proactive 
measures against accidents caused by fog and smog. 
Key Words: Computer Vision, Jetson Nano, Sensor Fusion and Switching, ADAS, Zero Visibilit 
Detection 
C JI...__,......_., 
111911 ... IIIIIIIIIIUCI MI 
Technology Used: 
DL model, Python, OpenCV. 
Supervisor Name: 
Dr. Muhammad Tariq 
Group Members: 
Abdullah Ali (i22 - 2153} 
Ali Shahid (i22 - 2199) 
Zain Khalid (i22 - 2161) 
FAST NUCES ISLAMABAD CAMPUS