Early Smoke Detection Based on Improved YOLO-PCA Network
Early detection of smoke having indistinguishable pixel intensities in digital images is a difficult task. To better maintain fire surveillance, early smoke detection is crucial. To solve the problem, we have integrated the principal component analysis (PCA) as a pre-processing module with the impro...
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-03-01
|
| Series: | Fire |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2571-6255/5/2/40 |
| _version_ | 1797434675387432960 |
|---|---|
| author | Muhammad Masoom S. Qixing Zhang Peiwen Dai Yang Jia Yongming Zhang Jiping Zhu Jinjun Wang |
| author_facet | Muhammad Masoom S. Qixing Zhang Peiwen Dai Yang Jia Yongming Zhang Jiping Zhu Jinjun Wang |
| author_sort | Muhammad Masoom S. |
| collection | DOAJ |
| description | Early detection of smoke having indistinguishable pixel intensities in digital images is a difficult task. To better maintain fire surveillance, early smoke detection is crucial. To solve the problem, we have integrated the principal component analysis (PCA) as a pre-processing module with the improved version of You Only Look Once (YOLOv3). The ordinary YOLOv3 structure has been improved after inserting one extra detection scale at stride-4 specifically to detect immense small smoke instances in the wild. The improved network design establishes a sequential relation between feature maps of lower spatial information and fine-grained semantic information in up-sampled maps via skip connections and concatenation operations. The testing of the improved model is carried out on self-prepared smoke datasets. In digital images, the smoke instances are captured in various complicated environments, for example, the mountains and fog in the background. A principal component analysis (PCA) helps in useful features selection and abandons the involvement of redundant features in the testing of the trained network hence, overcoming the latency at inference stage. In addition, to process small smoke images as positive samples during training, new sizes of anchors are calculated on small smoke data at a specified Intersection over Union (IoU) threshold. The experimental results show the improvement in precision rate, recall rate, and mean harmonic (F1-score) by 2.67, 3.06, and 5.59 percentages. The respective improvements in average precision (AP) and mean average precision (mAP) are 1.66 and 2.78 percentages. |
| first_indexed | 2024-03-09T10:36:58Z |
| format | Article |
| id | doaj.art-214e935680374b048de9854f1293bbef |
| institution | Directory Open Access Journal |
| issn | 2571-6255 |
| language | English |
| last_indexed | 2024-03-09T10:36:58Z |
| publishDate | 2022-03-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Fire |
| spelling | doaj.art-214e935680374b048de9854f1293bbef2023-12-01T20:52:29ZengMDPI AGFire2571-62552022-03-01524010.3390/fire5020040Early Smoke Detection Based on Improved YOLO-PCA NetworkMuhammad Masoom S.0Qixing Zhang1Peiwen Dai2Yang Jia3Yongming Zhang4Jiping Zhu5Jinjun Wang6State Key Laboratory of Fire Sciences, University of Science and Technology of China, No. 96 Jinzhai Road, Hefei 230026, ChinaState Key Laboratory of Fire Sciences, University of Science and Technology of China, No. 96 Jinzhai Road, Hefei 230026, ChinaState Key Laboratory of Fire Sciences, University of Science and Technology of China, No. 96 Jinzhai Road, Hefei 230026, ChinaShaanxi Key Laboratory of Network Data Intelligent Processing, Xi’an University of Posts and Telecommunications, Xi’an 710121, ChinaState Key Laboratory of Fire Sciences, University of Science and Technology of China, No. 96 Jinzhai Road, Hefei 230026, ChinaState Key Laboratory of Fire Sciences, University of Science and Technology of China, No. 96 Jinzhai Road, Hefei 230026, ChinaState Key Laboratory of Fire Sciences, University of Science and Technology of China, No. 96 Jinzhai Road, Hefei 230026, ChinaEarly detection of smoke having indistinguishable pixel intensities in digital images is a difficult task. To better maintain fire surveillance, early smoke detection is crucial. To solve the problem, we have integrated the principal component analysis (PCA) as a pre-processing module with the improved version of You Only Look Once (YOLOv3). The ordinary YOLOv3 structure has been improved after inserting one extra detection scale at stride-4 specifically to detect immense small smoke instances in the wild. The improved network design establishes a sequential relation between feature maps of lower spatial information and fine-grained semantic information in up-sampled maps via skip connections and concatenation operations. The testing of the improved model is carried out on self-prepared smoke datasets. In digital images, the smoke instances are captured in various complicated environments, for example, the mountains and fog in the background. A principal component analysis (PCA) helps in useful features selection and abandons the involvement of redundant features in the testing of the trained network hence, overcoming the latency at inference stage. In addition, to process small smoke images as positive samples during training, new sizes of anchors are calculated on small smoke data at a specified Intersection over Union (IoU) threshold. The experimental results show the improvement in precision rate, recall rate, and mean harmonic (F1-score) by 2.67, 3.06, and 5.59 percentages. The respective improvements in average precision (AP) and mean average precision (mAP) are 1.66 and 2.78 percentages.https://www.mdpi.com/2571-6255/5/2/40improved YOLOv3principal component analysisearly smoke detectionimages pre-processing |
| spellingShingle | Muhammad Masoom S. Qixing Zhang Peiwen Dai Yang Jia Yongming Zhang Jiping Zhu Jinjun Wang Early Smoke Detection Based on Improved YOLO-PCA Network Fire improved YOLOv3 principal component analysis early smoke detection images pre-processing |
| title | Early Smoke Detection Based on Improved YOLO-PCA Network |
| title_full | Early Smoke Detection Based on Improved YOLO-PCA Network |
| title_fullStr | Early Smoke Detection Based on Improved YOLO-PCA Network |
| title_full_unstemmed | Early Smoke Detection Based on Improved YOLO-PCA Network |
| title_short | Early Smoke Detection Based on Improved YOLO-PCA Network |
| title_sort | early smoke detection based on improved yolo pca network |
| topic | improved YOLOv3 principal component analysis early smoke detection images pre-processing |
| url | https://www.mdpi.com/2571-6255/5/2/40 |
| work_keys_str_mv | AT muhammadmasooms earlysmokedetectionbasedonimprovedyolopcanetwork AT qixingzhang earlysmokedetectionbasedonimprovedyolopcanetwork AT peiwendai earlysmokedetectionbasedonimprovedyolopcanetwork AT yangjia earlysmokedetectionbasedonimprovedyolopcanetwork AT yongmingzhang earlysmokedetectionbasedonimprovedyolopcanetwork AT jipingzhu earlysmokedetectionbasedonimprovedyolopcanetwork AT jinjunwang earlysmokedetectionbasedonimprovedyolopcanetwork |