Omni-Dimensional Dynamic Convolution Meets Bottleneck Transformer: A Novel Improved High Accuracy Forest Fire Smoke Detection Model

Qian, Jingjing; Lin, Ji; Dan, Bai; Xu, Runsheng; Lin, Haifeng

doi:10.3390/f14040838

Cited by 17 publications

(8 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…How to solve the problem of ensuring that there are no network differences after transplanting PC end algorithms to drone onboard devices is also a need to consider. [ (Qian et al, 2023;Matta et al, 2012;Woo et al, 2018;Li et al, 2019;Liu X. et al, 2020;Ge and Chen, 2020;Li C. et al, 2022;Zhang Y. et al, 2022;Ouyang and Yu, 2022;Huang et al, 2023)]…”

Section: Discussionmentioning

confidence: 99%

A bolt defect detection method for transmission lines based on improved YOLOv5

Zou,

Sun,

et al. 2024

Front. Energy Res.

View full text Add to dashboard Cite

To solve the problem of bolt defects in unmanned aerial vehicle inspection that are difficult to identify quickly and accurately, this paper proposes a defect detection method based on the improved YOLOv5 anchor mechanism. Firstly, the Normalized Wasserstein distance (NWD) evaluation metric and the Intersection over Union evaluation metric are combined, and the experiment determines the appropriate weight for this combination. This way, the sensitivity of using IoU alone to small objecet detection anchor box threshold changes was reduced. Furthermore, Convolutional Block Attention Module is included into the head network architecture of yolov5 in order to prioritize significant information and suppress irrelevant features. Omni-dimensional Dynamic Convolution (ODConv) is used to replace convolution in MobileNetv2. The combination module is used as the new backbone of the YOLOv5 model. It simultaneously enhances the model’s capability to extract bolt defect object information, minimizes calculation requirements, and achieves lightweight detection across the entire model. Compared with the original algorithm, the model detection Accuracy Precision (AP) is increased by 30.1%, the mean Accuracy Precision is increased by 30.4%. Other evaluation metrics of the model, such as GFlOPs and Parameters, all decreased slightly. The above results show that the improved algorithm proposed in this paper greatly improves the detection accuracy of the model on the premise of ensuring that the model is as small as possible.

show abstract

Section: Discussionmentioning

confidence: 99%

A bolt defect detection method for transmission lines based on improved YOLOv5

Zou,

Sun,

et al. 2024

Front. Energy Res.

View full text Add to dashboard Cite

show abstract

“…Additionally, they employed swin-TPH as the detection head, utilizing its hierarchical structure to enhance the algorithm's ability to detect small smoke targets. Qian et al [39] introduced the omni-dimensional dynamic convolution and bottleneck transformer structures into the YOLOv5 backbone. This allowed the algorithm to pay more attention to global features in the feature extraction process.…”

Section: Smoke Detectionmentioning

confidence: 99%

FuF-Det: An Early Forest Fire Detection Method under Fog

Pang,

Wu,

Yuan

2023

Remote Sensing

View full text Add to dashboard Cite

In recent years, frequent forest fires have seriously threatened the earth’s ecosystem and people’s lives and safety. With the development of machine vision and unmanned aerial vehicle (UAVs) technology, UAV monitoring combined with machine vision has become an important development trend in forest fire monitoring. In the early stages, fire shows the characteristics of a small fire target and obvious smoke. However, the presence of fog interference in the forest will reduce the accuracy of fire point location and smoke identification. Therefore, an anchor-free target detection algorithm called FuF-Det based on an encoder–decoder structure is proposed to accurately detect early fire points obscured by fog. The residual efficient channel attention block (RECAB) is designed as a decoder unit to improve the problem of the loss of fire point characteristics under fog caused by upsampling. Moreover, the attention-based adaptive fusion residual module (AAFRM) is used to self-enhance the encoder features, so that the features retain more fire point location information. Finally, coordinate attention (CA) is introduced to the detection head to make the image features correspond to the position information, and improve the accuracy of the algorithm to locate the fire point. The experimental results show that compared with eight mainstream target detection algorithms, FuF-Det has higher average precision and recall as an early forest fire detection method in fog and provides a new solution for the application of machine vision to early forest fire detection.

show abstract

“…On the one hand, Liu et al [16] utilized the YOLOv5n model for forest fire detection, which can be easily deployed on low-power devices but fails to meet the accuracy requirements. On the other hand, Qian et al [17] introduced the OBDS model, which combines CNN and Transformer to extract global feature information from forest fire smoke images. Li et al [18] replaced the SPPF module with RFB in YOLOv5 to enable better focus on the global information of various forest fire and smoke.…”

Section: Introductionmentioning

confidence: 99%

FFYOLO: A Lightweight Forest Fire Detection Model Based on YOLOv8

Yun,

Zheng,

Lin

et al. 2024

Fire

View full text Add to dashboard Cite

Forest is an important resource for human survival, and forest fires are a serious threat to forest protection. Therefore, the early detection of fire and smoke is particularly important. Based on the manually set feature extraction method, the detection accuracy of the machine learning forest fire detection method is limited, and it is unable to deal with complex scenes. Meanwhile, most deep learning methods are difficult to deploy due to high computational costs. To address these issues, this paper proposes a lightweight forest fire detection model based on YOLOv8 (FFYOLO). Firstly, in order to better extract the features of fire and smoke, a channel prior dilatation attention module (CPDA) is proposed. Secondly, the mixed-classification detection head (MCDH), a new detection head, is designed. Furthermore, MPDIoU is introduced to enhance the regression and classification accuracy of the model. Then, in the Neck section, a lightweight GSConv module is applied to reduce parameters while maintaining model accuracy. Finally, the knowledge distillation strategy is used during training stage to enhance the generalization ability of the model and reduce the false detection. Experimental outcomes demonstrate that, in comparison to the original model, FFYOLO realizes an mAP0.5 of 88.8% on a custom forest fire dataset, which is 3.4% better than the original model, with 25.3% lower parameters and 9.3% higher frames per second (FPS).

show abstract

Omni-Dimensional Dynamic Convolution Meets Bottleneck Transformer: A Novel Improved High Accuracy Forest Fire Smoke Detection Model

Cited by 17 publications

References 33 publications

A bolt defect detection method for transmission lines based on improved YOLOv5

A bolt defect detection method for transmission lines based on improved YOLOv5

FuF-Det: An Early Forest Fire Detection Method under Fog

FFYOLO: A Lightweight Forest Fire Detection Model Based on YOLOv8

Contact Info

Product

Resources

About