IMPROVEMENT THE FIRE DETECTION PERFORMANCE USING COMPUTER VISION BASED ON YOLOV5 NEURAL NETWORK

Abstract

Early and accurate in fire detection is an urgent requirement for fire alarm
systems. To contribute to solving this problem, this paper presents a fire
detection model based on YOLOv5 convolutional neural network (CNNs). Using
YOLOv5 convolution neural network to fire detection has given high accuracy
upper 90% and real time detection. The current fire and explosion situation
combine analysis of the operation of fire alarm systems be used, a fire detection
based on YOLOv5 convolutiona l neural network when combine a sercurity
camera systerm improve the performance of the fire alarm system.