Flame and Smoke Detection Algorithm Based on ODConvBS-YOLOv5s

Abstract

Real-time and accurate detection of flame and smoke is an important prerequisite to reduce the loss caused by fire. There exists some problems in traditional flame and smoke detection algorithm, such as low accuracy, high miss rate, low detection efficiency and low detection rate of small targets. This paper proposes a YOLOv5s flame smoke detection algorithm based on ODConvBS. Firstly, in the YOLOv5s backbone network, the ordinary convolution block is replaced by ODConvBS to realize the extraction of attention features of the convolution kernel; Secondly, Gnconv is introduced into Neck to improve the model's high-order spatial information extraction ability; then the Shuffle Attention module is added at the end of Neck to facilitate the fusion of different groups of features; At last, in the prediction section, the SIOU loss function, which can account for the angle of the prediction frame vector, is utilized to speed up model convergence. When utilizing the self-made flame and smoke data set, the upgraded YOLOv5s model mAP grew by 9.3% .At the same time, the accuracy rate and the recall rate and the detection speed increased to 83.5%, 83.7%, 33.3FPS respectively.

AE-YOLOv5 for Detection of Power Line Insulator Defects

“…Since it was proposed, it has undergone several model iterations, forming the YOLO model family. With its flexibility and excellent performance, YOLOv5 model has been widely used in traffic sign recognition [21], pedestrian detection [22], automatic driving [23], fire detection and alarm [13] and other tasks. It can be deployed to kinds of embedded hardware platforms because of its low computational cost.…”

“…The objective of this research is to explore a novel model capable of real-time and accurate detection of insulator defects. Noting the rapid development of YOLO models [11] in the field of general object detection task and their wide application in various industries [12,13], this paper introduces and enhances YOLOv5 model to make it adapt to the task of insulator defect detection in complex environments. Different from Faster R-CNN, YOLO innovatively proposes a single-stage object detection method, which achieves a qualitative breakthrough in computing speed and is often used for various real-time object detection tasks.…”

Steel Surface Defect Detection Using Improved Deep Learning Algorithm: ECA-SimSPPF-SIoU-Yolov5

“…However, further improvements in detection accuracy are still necessary for real-world applications. Ma et al [9] replaced the convolutional layers in YOLOv5 and applied it to fire and smoke detection scenarios, achieving a detection accuracy of 87.6%. However, since it was applied to fire and smoke detection, further improvements in detection accuracy are required for practical real-life applications.…”

A Flame Detection Algorithm Based on Improved YOLOv7

“…Following that, the YOLO series has been developed over many upgraded versions. There are many flame detection algorithms based on YOLOv3, YOLOv4, and YOLOv5 [5][6][7][8][9]. YOLOv7 [15] is the latest version of the YOLO series, and in [16], it is proved that the performance of YOLOv7 is obviously better than YOLOv5 and YOLOv6 in the aspect of target detection.…”

“…Convolutional neural networks have strong learning ability, fault tolerance, and fast speed; thus, they are commonly used in image recognition and classification. Currently, the convolutional neural networks (CNNs) used for object detection mainly include region-convolutional neural networks (R-CNN) [4] and YOLO series [5][6][7][8][9]. Compared with other convolutional neural networks, the YOLO series can better extract global information from images and can be trained end-to-end, which assures them as a more suitable option for flame detection.…”