7 research outputs found
An Order and Difference Local Binary Pattern for Hyperspectral Texture Classification
No full textInternational audienceThis paper proposes a new hyperspectral texture descriptor, which is a variant of Local Binary Pattern (LBP) for hyperspectral imaging. This descriptor effectively describes the texture information in hyperspectral images while addressing the issue of high dimensionality. The proposed descriptor enhances the LBP by using an ordering approach along the spectral dimension to capture local neighboring relationships across different spectral bands. It then approximates the second-order derivatives by calculating differences between adjacent ordered bands to capture the ordered spectral variation. Finally, it applies rotation invariant uniform LBP separately on each ordered and differenced band. The proposed descriptor, Order Difference Local Binary Pattern (ODLBP), concatenates the resulting histograms. It is tested on several hyperspectral image datasets for texture classification. The results demonstrate that the proposed ODLBP has comparable performance with the most discriminative LBP descriptors while maintaining a reduced dimensionality space, indicating its effectiveness in hyperspectral texture analysis
A lightweight spatial and spectral CNN model for classifying floating marine plastic debris using hyperspectral images
No full textInternational audienceMarine plastic debris poses a significant environmental threat. In order to study and combat this pollution, efficient and automated detection methods are essential. Hyperspectral imaging and deep learning provide a robust framework for classifying floating marine plastic debris. However, deep learning approaches often suffer from high computational complexity and limited interpretability. In addition, hyperspectral images are high-dimensional data that must be analyzed efficiently. To overcome these limitations, this paper proposes the Lightweight Spatial and Spectral Hyperspectral CNN (LSS-HCNN), a deep learning model designed to enhance classification accuracy while improving computational efficiency and interpretability. The proposed model first applies spatialwise convolutions to extract spatial features from individual spectral bands, then uses spectralwise convolutions to extract relationships between spectral bands. Additionally, a Squeeze-and-Excitation (SE) block improves interpretability by focusing on the most informative spectral bands. Experiments were conducted on three hyperspectral datasets containing various materials and four dedicated floating plastic datasets, including a new plastic waste dataset. It provides images in three spectral configurations: visible-near-infrared (VIS-NIR), near-infrared-shortwave-infrared (NIR-SWIR), and a fused domain. Results show that LSS-HCNN outperforms traditional handcrafted descriptors and deep-learning models, particularly for floating marine plastics. It achieves a mean classification accuracy of 97.64% while reducing model complexity. Compared to standard 2D-CNN, it reduces the number of parameters by over 80% and Floating Point Operations (FLOPS) by a factor of 7. Moreover, SE block analysis reveals that NIR-SWIR bands contribute the most to plastic classification. This highlights LSS-HCNN as an efficient marine plastic debris classification model, supporting environmental monitoring efforts
Système d'imagerie hyperspectrale embarqué sur un véhicule téléopéré aquatique pour la détection et l'identification automatique de déchets marins flottants
No full textNational audienc
Benchmark dataset and classification of marine plastic waste acquired by a remote hyperspectral imaging system embedded on an aquatic drone.
No full textInternational audienceThe problem of marine debris, particularly plastic waste, presents significant challenges for environmental pollution. This work introduces a new benchmark dataset specifically designed for the classification of floating plastic litter and captured by a remote hyperspectral imaging system embedded on an unmanned aquatic drone. This innovative system is equipped with a hyperspectral camera that covers the spectral range between 900 nm and 1700 nm, from near-infrared (NIR) to short-wave infrared (SWIR). The acquired dataset primarily focuses on various types of marine plastic debris collected from the coastline of the Côte d’Opale in France, containing seven plastic types commonly found in marine environments, in addition to a category for non-plastic debris. The hyperspectral images of this marine waste were acquired in the wave and current flume tank of IFREMER in Boulognesur-Mer. This facility is able to precisely control the speed of the water on which the debris floats and to reproduce different situations encountered in real conditions. To analyze the data, we applied several standard machine learning algorithms (Nearest Neighbor, Support Vector Machine, Random Forest, Logistic Regression, Naïve Bayes, and Linear Discriminant Analysis) as benchmarks to classify and identify debris types based on their spectral signatures. In addition, Principal Component Analysis (PCA) was applied to reduce the dimensionality of the representation space. The dataset and experimental results provide valuable resources for researchers aiming to improve marine debris detection and identification techniques using hyperspectral imaging technology
Benchmark dataset and classification of marine plastic waste acquired by a remote hyperspectral imaging system embedded on an aquatic drone.
No full textInternational audienceThe problem of marine debris, particularly plastic waste, presents significant challenges for environmental pollution. This work introduces a new benchmark dataset specifically designed for the classification of floating plastic litter and captured by a remote hyperspectral imaging system embedded on an unmanned aquatic drone. This innovative system is equipped with a hyperspectral camera that covers the spectral range between 900 nm and 1700 nm, from near-infrared (NIR) to short-wave infrared (SWIR). The acquired dataset primarily focuses on various types of marine plastic debris collected from the coastline of the Côte d’Opale in France, containing seven plastic types commonly found in marine environments, in addition to a category for non-plastic debris. The hyperspectral images of this marine waste were acquired in the wave and current flume tank of IFREMER in Boulognesur-Mer. This facility is able to precisely control the speed of the water on which the debris floats and to reproduce different situations encountered in real conditions. To analyze the data, we applied several standard machine learning algorithms (Nearest Neighbor, Support Vector Machine, Random Forest, Logistic Regression, Naïve Bayes, and Linear Discriminant Analysis) as benchmarks to classify and identify debris types based on their spectral signatures. In addition, Principal Component Analysis (PCA) was applied to reduce the dimensionality of the representation space. The dataset and experimental results provide valuable resources for researchers aiming to improve marine debris detection and identification techniques using hyperspectral imaging technology
Benchmark dataset and classification of marine plastic waste acquired by a remote hyperspectral imaging system embedded on an aquatic drone.
No full textInternational audienceThe problem of marine debris, particularly plastic waste, presents significant challenges for environmental pollution. This work introduces a new benchmark dataset specifically designed for the classification of floating plastic litter and captured by a remote hyperspectral imaging system embedded on an unmanned aquatic drone. This innovative system is equipped with a hyperspectral camera that covers the spectral range between 900 nm and 1700 nm, from near-infrared (NIR) to short-wave infrared (SWIR). The acquired dataset primarily focuses on various types of marine plastic debris collected from the coastline of the Côte d’Opale in France, containing seven plastic types commonly found in marine environments, in addition to a category for non-plastic debris. The hyperspectral images of this marine waste were acquired in the wave and current flume tank of IFREMER in Boulognesur-Mer. This facility is able to precisely control the speed of the water on which the debris floats and to reproduce different situations encountered in real conditions. To analyze the data, we applied several standard machine learning algorithms (Nearest Neighbor, Support Vector Machine, Random Forest, Logistic Regression, Naïve Bayes, and Linear Discriminant Analysis) as benchmarks to classify and identify debris types based on their spectral signatures. In addition, Principal Component Analysis (PCA) was applied to reduce the dimensionality of the representation space. The dataset and experimental results provide valuable resources for researchers aiming to improve marine debris detection and identification techniques using hyperspectral imaging technology
Benchmark dataset and classification of marine plastic waste acquired by a remote hyperspectral imaging system embedded on an aquatic drone.
No full textInternational audienceThe problem of marine debris, particularly plastic waste, presents significant challenges for environmental pollution. This work introduces a new benchmark dataset specifically designed for the classification of floating plastic litter and captured by a remote hyperspectral imaging system embedded on an unmanned aquatic drone. This innovative system is equipped with a hyperspectral camera that covers the spectral range between 900 nm and 1700 nm, from near-infrared (NIR) to short-wave infrared (SWIR). The acquired dataset primarily focuses on various types of marine plastic debris collected from the coastline of the Côte d’Opale in France, containing seven plastic types commonly found in marine environments, in addition to a category for non-plastic debris. The hyperspectral images of this marine waste were acquired in the wave and current flume tank of IFREMER in Boulognesur-Mer. This facility is able to precisely control the speed of the water on which the debris floats and to reproduce different situations encountered in real conditions. To analyze the data, we applied several standard machine learning algorithms (Nearest Neighbor, Support Vector Machine, Random Forest, Logistic Regression, Naïve Bayes, and Linear Discriminant Analysis) as benchmarks to classify and identify debris types based on their spectral signatures. In addition, Principal Component Analysis (PCA) was applied to reduce the dimensionality of the representation space. The dataset and experimental results provide valuable resources for researchers aiming to improve marine debris detection and identification techniques using hyperspectral imaging technology
