A high resolution burned area detector for Sentinel-2 and Landsat-8

Zanetti, Massimo; Marinelli, Daniele; Bertoluzza, Manuel; Saha, Sudipan; Bovolo, Francesca; Bruzzone, Lorenzo; Magliozzi, M.L.; Zavagli, Massimo; Costantini, E.

doi:10.1109/multi-temp.2019.8866958

Cited by 8 publications

(6 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Another research study [33] attempted to integrate multiple SIs: After performing an SI separability analysis, SI values were reduced to a common domain, converted to positive and negative evidence [38], and integrated using fuzzy membership functions; finally, burned area pixels were identified, applying different operators, such as ordered weighted averaging [51]. Alternatively, research studies using multiple SIs adopted a set of empirical conditional statements, also defined as decision rules, to identify burned area maps [31] even with the support of a temporal harmonizer algorithm [64], or use distribution-modeling algorithms trained by hyperspectral indexes and hotspots to model and map burned areas [47]. Compared to other methodologies described in the literature, the approach presented in this study has the strengths of significantly reducing the use of pre-defined set of parameters, thanks to the adaptive thresholding, and rejecting the use of SIs whose mapping performances are not adequate for specific site conditions, thanks to the AIX.…”

Section: Discussionmentioning

confidence: 99%

Agreement Index for Burned Area Mapping: Integration of Multiple Spectral Indices Using Sentinel-2 Satellite Images

et al. 2020

View full text Add to dashboard Cite

Identifying fire-affected areas is of key importance to support post-fire management strategies and account for the environmental impact of fires. The availability of high spatial and temporal resolution optical satellite data enables the development of procedures for detailed and prompt post-fire mapping. This study proposes a novel approach for integrating multiple spectral indices to generate more accurate burned area maps by exploiting Sentinel-2 images. This approach aims to develop a procedure to combine multiple spectral indices using an adaptive thresholding method and proposes an agreement index to map the burned areas by optimizing omission and commission errors. The approach has been tested for the burned area classification of four study areas in Italy. The proposed agreement index combines multiple spectral indices to select the actual burned pixels, to balance the omission and commission errors, and to optimize the overall accuracy. The results showed the spectral indices singularly performed differently in the four study areas and that high levels of commission errors were achieved, especially for wildfires which occurred during the fall season (up to 0.93) Furthermore, the agreement index showed a good level of accuracy (minimum 0.65, maximum 0.96) for all the study areas, improving the performance compared to assessing the indices individually. This suggests the possibility of testing the methodology on a large set of wildfire cases in different environmental conditions to support the decision-making process. Exploiting the high resolution of optical satellite data, this work contributes to improving the production of detailed burned area maps, which could be integrated into operational services based on the use of Earth Observation products for burned area mapping to support the decision-making process.

show abstract

Section: Discussionmentioning

confidence: 99%

Agreement Index for Burned Area Mapping: Integration of Multiple Spectral Indices Using Sentinel-2 Satellite Images

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Then, the temporal harmonization method was used to confirm the burned areas and reduce the false alarms. Experimental results proved the ability of this method to correctly detect the burned areas and reduce false alarms [84].…”

Section: Post-fire Mapping Based Satellite Remote Sensing Imagerymentioning

confidence: 80%

“…Zanetti et al [84] proposed a multitemporal automatic and unsupervised method to detect burned areas using image time series acquired by Sentinel-2 and Landsat-8. At first, the normalized fire index method detected the candidate burned areas in each input image.…”

Section: Post-fire Mapping Based Satellite Remote Sensing Imagerymentioning

confidence: 99%

Wildfire Segmentation Using Deep Vision Transformers

Ghali

Akhloufi

Jmal³

et al. 2021

Remote Sensing

View full text Add to dashboard Cite

In this paper, we address the problem of forest fires’ early detection and segmentation in order to predict their spread and help with fire fighting. Techniques based on Convolutional Networks are the most used and have proven to be efficient at solving such a problem. However, they remain limited in modeling the long-range relationship between objects in the image, due to the intrinsic locality of convolution operators. In order to overcome this drawback, Transformers, designed for sequence-to-sequence prediction, have emerged as alternative architectures. They have recently been used to determine the global dependencies between input and output sequences using the self-attention mechanism. In this context, we present in this work the very first study, which explores the potential of vision Transformers in the context of forest fire segmentation. Two vision-based Transformers are used, TransUNet and MedT. Thus, we design two frameworks based on the former image Transformers adapted to our complex, non-structured environment, which we evaluate using varying backbones and we optimize for forest fires’ segmentation. Extensive evaluations of both frameworks revealed a performance superior to current methods. The proposed approaches achieved a state-of-the-art performance with an F1-score of 97.7% for TransUNet architecture and 96.0% for MedT architecture. The analysis of the results showed that these models reduce fire pixels mis-classifications thanks to the extraction of both global and local features, which provide finer detection of the fire’s shape.

show abstract

“…As shown in Figure 2, the spectral bands presenting noticeable correlations (medium or high) with both the dNBR index and the target variable GT (i.e., the damage severity level) are B06, B07, B08, B8A, and B09. Except for B09, the other bands are known in the literature for the computation of the most used indexes for fire detection, such as the Burned Area Index (BAI) [17], the Burned Area Index for Sentinel2 (BAIS2) [38], the Normalized Burn Ratio (NBR) [39], the Normalized Burn Ratio 2 (NBR2) [16], and the Mid-Infrared Burned Index (MIRBI) [40]. Moreover, B07 and B08 are used for the computation of (i) the vegetation index, Normalized Difference Vegetation Index (NDVI) [41], and (ii) the water index, Normalized Difference Water Index (NDWI) [42], respectively.…”

Section: Data Acquisition Preprocessing and Analysismentioning

confidence: 99%

Double-Step U-Net: A Deep Learning-Based Approach for the Estimation of Wildfire Damage Severity through Sentinel-2 Satellite Data

2020

View full text Add to dashboard Cite

Wildfire damage severity census is a crucial activity for estimating monetary losses and for planning a prompt restoration of the affected areas. It consists in assigning, after a wildfire, a numerical damage/severity level, between 0 and 4, to each sub-area of the hit area. While burned area identification has been automatized by means of machine learning algorithms, the wildfire damage severity census operation is usually still performed manually and requires a significant effort of domain experts through the analysis of imagery and, sometimes, on-site missions. In this paper, we propose a novel supervised learning approach for the automatic estimation of the damage/severity level of the hit areas after the wildfire extinction. Specifically, the proposed approach, leveraging on the combination of a classification algorithm and a regression one, predicts the damage/severity level of the sub-areas of the area under analysis by processing a single post-fire satellite acquisition. Our approach has been validated in five different European countries and on 21 wildfires. It has proved to be robust for the application in several geographical contexts presenting similar geological aspects.

show abstract

A high resolution burned area detector for Sentinel-2 and Landsat-8

Cited by 8 publications

References 10 publications

Agreement Index for Burned Area Mapping: Integration of Multiple Spectral Indices Using Sentinel-2 Satellite Images

Agreement Index for Burned Area Mapping: Integration of Multiple Spectral Indices Using Sentinel-2 Satellite Images

Wildfire Segmentation Using Deep Vision Transformers

Double-Step U-Net: A Deep Learning-Based Approach for the Estimation of Wildfire Damage Severity through Sentinel-2 Satellite Data

Contact Info

Product

Resources

About