Smoke Image Segmentation Based on Color Model

Deng, Xing; Yu, Zhongming; Wang, Lin; Li, Jinlan

doi:10.24212/2179-3565.2015v6i2p130-138

Cited by 10 publications

(3 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although current semantic segmentation methods (Long, Shelhamer, and Darrell 2015;Lin et al 2017;Chen et al 2018;Strudel et al 2021;Xie et al 2021;Cheng et al 2022) are effective at segmenting regular objects (i.e., those with clear outlines and roughly the same shape), these generic algorithms are not suitable for early smoke segmentation due to its varying transparency and small scale. Traditional smoke segmentation methods have mainly focused on extracting high-quality color and texture features (Mahmoud and Ren 2019;Xing et al 2015;Yuan, Liu, and Zhang 2019), and deep-learning-based methods (Yuan et al 2022(Yuan et al , 2019aJing, Meng, and Hou 2023) mainly focus on extracting features with larger receptive fields to handle the variability and blurred edges of smoke. However, these methods suffer from high miss detection rate and low precision in segmenting small and transparent early smoke.…”

Section: Related Workmentioning

confidence: 99%

FoSp: Focus and Separation Network for Early Smoke Segmentation

Yao,

Zhao,

Peng

et al. 2024

AAAI

View full text Add to dashboard Cite

Early smoke segmentation (ESS) enables the accurate identification of smoke sources, facilitating the prompt extinguishing of fires and preventing large-scale gas leaks. But ESS poses greater challenges than conventional object and regular smoke segmentation due to its small scale and transparent appearance, which can result in high miss detection rate and low precision. To address these issues, a Focus and Separation Network (FoSp) is proposed. We first introduce a Focus module employing bidirectional cascade which guides low-resolution and high-resolution features towards mid-resolution to locate and determine the scope of smoke, reducing the miss detection rate. Next, we propose a Separation module that separates smoke images into a pure smoke foreground and a smoke-free background, enhancing the contrast between smoke and background fundamentally, improving segmentation precision. Finally, a Domain Fusion module is developed to integrate the distinctive features of the two modules which can balance recall and precision to achieve high F_beta. Futhermore, to promote the development of ESS, we introduce a high-quality real-world dataset called SmokeSeg, which contains more small and transparent smoke than the existing datasets. Experimental results show that our model achieves the best performance on three available smoke segmentation datasets: SYN70K (mIoU: 83.00%), SMOKE5K (F_beta: 81.6%) and SmokeSeg (F_beta: 72.05%). The code can be found at https://github.com/LujianYao/FoSp.

show abstract

Section: Related Workmentioning

confidence: 99%

FoSp: Focus and Separation Network for Early Smoke Segmentation

Yao,

Zhao,

Peng

et al. 2024

AAAI

View full text Add to dashboard Cite

show abstract

“…Although this method could segment smoke regions, the integrity of the main body of smoke was not strong, and the segmentation performance of the overall shape and edge regions of smoke was poor. Deng Xing et al [24] proposed combining the results of K-means clustering in the Lab color space and threshold segmentation in the HSV color space, followed by filtering and region labeling for noise removal. This method could segment the main body of smoke, but some non-smoke regions were still segmented.…”

Section: Smoke Segmentationmentioning

confidence: 99%

U2-Net_S Fire Smoke Segmentation Based on Multiscale Edge Fusion Convolutional Encoder and Edge Supervision

Zheng,

Wang,

Gan

et al. 2023

Preprint

View full text Add to dashboard Cite

The production of smoke is a sign of fire occurrence, and in the monitoring and prevention of disasters, the research value of smoke is higher than that of flames. Segmentation of smoke targets is of great significance for the study of smoke characteristics. However, traditional methods for smoke target segmentation are limited in accuracy and are easily affected by the scene. Existing smoke segmentation models rely on a large-scale classification backbone, which sacrifices high-resolution feature mappings to achieve deeper architectures. These models neglect the relatively important local details and global contrast information that are crucial for smoke segmentation, and their segmentation results for smoke edge details are also not satisfactory. In this paper, the Sobel operator and multi-scale convolutional encoder are fused for feature extraction, and the residual mixing attention mechanism as well as the edge segmentation module are added to the model, and a multi-scale segmentation model for fire smoke edge segmentation is proposed, which improves the segmentation accuracy, segmentation precision and edge segmentation. The test results of the homemade smoke segmentation dataset show that the model's mIoU can reach up to 90.48% and finally stabilized at 90.09%, mPA is up to 96.96%, MAE is reduced to 0.02, and MaxF1 reaches 94.37%, which is superior to the current semantic segmentation model. Compared with the base model U2-Net improves 3.26% (AVG20) and 2.68% (TOP1) in mIoU metrics, and 0.95% and 1.36% in mPA and MaxF1 metrics, respectively. The method proposed in this paper is more suitable for fire smoke segmentation, which is important for smoke characterization and smoke source point identification research.

show abstract

“…Haridasan et al [47] applied multiple color space transforms to RGB fire images with flame areas and fused the resulting features using concatenation to improve classification performance. Xing D et al [48] achieved refined smoke segmentation results by merging smoke segmentation regions using HSV and LAB color spaces. Similarly, Prema E et al [49] and Pundir, A.S. et al [50] used color spaces (YUV and YCbCr, respectively) as color criteria and combined them with other feature extraction analyses to detect smoke presence in video frames.…”

Section: Introductionmentioning

confidence: 99%

Forest Fire Smoke Detection Based on Multiple Color Spaces Deep Feature Fusion

Han,

Tian,

Zheng

et al. 2024

Forests

View full text Add to dashboard Cite

The drastic increase of forest fire occurrence, which in recent years has posed severe threat and damage worldwide to the natural environment and human society, necessitates smoke detection of the early forest fire. First, a semantic segmentation method based on multiple color spaces feature fusion is put forward for forest fire smoke detection. Considering that smoke images in different color spaces may contain varied and distinctive smoke features which are beneficial for improving the detection ability of a model, the proposed model integrates the function of multi-scale and multi-type self-adaptive weighted feature fusion with attention augmentation to extract the enriched and complementary fused features of smoke, utilizing smoke images from multi-color spaces as inputs. Second, the model is trained and evaluated on part of the FIgLib dataset containing high-quality smoke images from watchtowers in the forests, incorporating various smoke types and complex background conditions, with a satisfactory smoke segmentation result for forest fire detection. Finally, the optimal color space combination and the fusion strategy for the model is determined through elaborate and extensive experiments with a superior segmentation result of 86.14 IoU of smoke obtained.

show abstract

Smoke Image Segmentation Based on Color Model

Cited by 10 publications

References 11 publications

FoSp: Focus and Separation Network for Early Smoke Segmentation

FoSp: Focus and Separation Network for Early Smoke Segmentation

U2-Net_S Fire Smoke Segmentation Based on Multiscale Edge Fusion Convolutional Encoder and Edge Supervision

Forest Fire Smoke Detection Based on Multiple Color Spaces Deep Feature Fusion

Contact Info

Product

Resources

About