Big Map R-CNN for object detection in large-scale remote sensing images

Sea fog detection (SFD) presents a significant challenge in the field of intelligent Earth observation, particularly in analyzing meteorological satellite imagery. Akin to various vision tasks, ImageNet pre-training is commonly used for pre-training SFD. However, in the context of multi-spectral meteorological satellite imagery, the initial step of deep learning has received limited attention. Recently, pre-training with Very High-Resolution (VHR) satellite imagery has gained increased popularity in remote-sensing vision tasks, showing the potential to replace ImageNet pre-training. However, it is worth noting that the meteorological satellite imagery applied in SFD, despite being an application of computer vision in remote sensing, differs greatly from VHR satellite imagery. To address the limitation of pre-training for SFD, this paper introduces a novel deep-learning paradigm to the meteorological domain driven by Masked Image Modeling (MIM). Our research reveals two key insights: (1) Pre-training with meteorological satellite imagery yields superior SFD performance compared to pre-training with nature imagery and VHR satellite imagery. (2) Incorporating the architectural characteristics of SFD models into a vanilla masked autoencoder (MAE) can augment the effectiveness of meteorological pre-training. To facilitate this research, we curate a pre-training dataset comprising 514,655 temporal multi-spectral meteorological satellite images, covering the Bohai Sea and Yellow Sea regions, which have the most sea fog occurrence. The longitude ranges from 115.00E to 128.75E, and the latitude ranges from 27.60N to 41.35N. Moreover, we introduce SeaMAE, a novel MAE that utilizes a Vision Transformer as the encoder and a convolutional hierarchical decoder, to learn meteorological representations. SeaMAE is pre-trained on this dataset and fine-tuned for SFD, resulting in state-of-the-art performance. For instance, using the ViT-Base as the backbone, SeaMAE pre-training which achieves 64.18% surpasses from-scratch learning, natural imagery pre-training, and VRH satellite imagery pre-training by 5.53%, 2.49%, and 2.21%, respectively, in terms of Intersection over Union of SFD.

This paper presents a two-level Active Learning (AL) classification method for the interactive detection of earthquake-induced debris via the synergetic use of post-disaster Very High Resolution (VHR) satellite and local decimeter-resolution aerial images. The proposed method is performed by interactively guiding the human expert in the collection of labeled training samples from aerial images and optimally planning further aerial surveys. A label propagation mechanism is adopted to reliably propagate the labeled pixels annotated by the user on the aerial image by visual photointerpretation to the (lower resolution) satellite image. The experimental analysis is carried out using post-disaster images of the 2010 Yushu earthquake in China. The obtained results confirm that the proposed method can significantly reduce the user annotation effort and the cost for acquiring additional aerial images leading to more accurate and timely debris detection maps.

Monitoring the variation of urban expansion is crucial for sustainable urban planning and cultural heritage management. This paper proposes an approach for the semantic segmentation of very-high-resolution (VHR) satellite imagery to detect the changes in urban sprawl in the surroundings of Chan Chan, a UNESCO World Heritage Site in Peru. This study explores the effectiveness of combining Mix Transformer encoders with U-Net architectures to improve feature extraction and spatial context understanding in VHR satellite imagery. The integration of active contour loss functions further enhances the model’s ability to delineate complex urban boundaries, addressing the challenges posed by the heterogeneous landscape surrounding the archaeological complex of Chan Chan. The results demonstrate that the proposed approach achieves accurate semantic segmentation on images of the study area from different years. Quantitative results showed that the U-Net-scse model with an MiTB5 encoder achieved the best performance with respect to SegFormer and FT-UNet-Former, with IoU scores of 0.8288 on OpenEarthMap and 0.6743 on Chan Chan images. Qualitative analysis revealed the model’s effectiveness in segmenting buildings across diverse urban and rural environments in Peru. Utilizing this approach for monitoring urban expansion over time can enable managers to make informed decisions aimed at preserving cultural heritage and promoting sustainable urban development.

Case instance segmentation of small farmland based on Mask R-CNN of feature pyramid network with double attention mechanism in high resolution satellite images

Pine wilt disease has caused severe damage to China’s forest ecosystems. Utilizing the rich information from very-high-resolution (VHR) satellite imagery for large-scale and accurate monitoring of pine wilt disease is a crucial approach to curbing its spread. However, current research on identifying infected trees using VHR satellite imagery and deep learning remains extremely limited. This study introduces several advanced self-attention algorithms into the task of satellite-based monitoring of pine wilt disease to enhance detection performance. We constructed a dataset of discolored pine trees affected by pine wilt disease using imagery from the Gaofen-2 and Gaofen-7 satellites. Within the unified semantic segmentation framework MMSegmentation, we implemented four single-head attention models—NLNet, CCNet, DANet, and GCNet—and two multi-head attention models—Swin Transformer and SegFormer—for the accurate semantic segmentation of infected trees. The model predictions were further analyzed through visualization. The results demonstrate that introducing appropriate self-attention algorithms significantly improves detection accuracy for pine wilt disease. Among the single-head attention models, DANet achieved the highest accuracy, reaching 73.35%. The multi-head attention models exhibited an excellent performance, with SegFormer-b2 achieving an accuracy of 76.39%, learning the features of discolored pine trees at the earliest stage and converging faster. The visualization of model inference results indicates that DANet, which integrates convolutional neural networks (CNNs) with self-attention mechanisms, achieved the highest overall accuracy at 94.43%. The use of self-attention algorithms enables models to extract more precise morphological features of discolored pine trees, enhancing user accuracy while potentially reducing production accuracy.

Oil palm trees detection and counting on Microsoft Bing Maps Very High Resolution (VHR) satellite imagery and Unmanned Aerial Vehicles (UAV) data using image processing thresholding approach

Reliable building delineation in very high resolution (VHR) satellite imagery can be achieved by precise disparity information extracted from stereo pairs. However, off-nadir VHR images over urban areas contain many occlusions due to building leaning that creates gaps in the extracted disparity maps. The typical approach to fill these gaps is by interpolation. However, it inevitably degrades the quality of the disparity map and reduces the accuracy of building detection. Thus, this research proposes a registration-based technique for mapping the disparity of off-terrain objects to avoid the need for disparity interpolation and normalization. The generated disparity by the proposed technique is then used to support building detection in off-nadir VHR satellite images. Experiments in a high-rise building area confirmed that 75 percent of the detected building roofs overlap precisely the reference data, with almost 100 percent correct detection. These accuracies are substantially higher than those achieved by other published research.

The paper focuses on the application of the Cross-Correlation Analysis (CCA) technique for quantifying changes of semi-natural grasslands to artificial structures at different spatial resolutions (grain) on a Mediterranean Natura 2000 site as a further test case of the results reported in [1]. In that work the CCA was applied to detect specific changes associated with agricultural intensification and fires. A semi-natural grasslands layer extracted from an existing Land Cover/Land Use map (1:5000, time T1) was considered as input to the CCA jointly with a Very High Resolution (VHR) WorldView-2 satellite image (2 meters spatial resolution, time T2) and with a High Resolution Landsat 8 OLS satellite image (30 meters spatial resolution, time T2), with T2 > T1, respectively, for the fine and the coarse scale analysis. The results were compared to those obtained by applying a traditional Post Classification Comparison technique to the same reference at time T1 map and an updated at time T2 map obtained by a knowledge driven classification of four multi-seasonal Worldview-2 input images, according to [1]. Also in this case the CCA technique results encouraging offering the possibility to reduce the costs of change detection when the acquisition of multi-seasonal VHR images at time T2 is too expensive or when no archive VHR image is available in the past for comparison between at time T1 and T2 images. The areas of change detected at VHR and HR were quite similar for the specific transition analyzed with larger error values in HR change images.

Pléiades satellite images for deriving forest metrics in the Alpine region

Abstract. The Very High Resolution (VHR) satellite imageries such us Pleiades, WorldView-2, GeoEye-1 used for precise mapping purpose must be corrected from any distortion to achieve the expected accuracy. Orthorectification is performed to eliminate geometric errors of the VHR satellite imageries. Orthorectification requires main input data such as Digital Elevation Model (DEM) and Ground Control Point (GCP). The VHR systematic-ortho imageries were generated using SRTM 30m DEM without using any GCP data. The accuracy value differences of VHR systematic-ortho imageries and VHR orthorectified imageries using GCP currently is not exactly defined. This study aimed to identified the accuracy comparison of VHR systematic-ortho imageries against orthorectified imageries using GCP. Orthorectified imageries using GCP created by using Rigorous model. Accuracy evaluation is calculated by using several independent check points.

As the acquisition of very high resolution (VHR) satellite images becomes easier owing to technological advancements, ever more stringent requirements are being imposed on automatic image interpretation. Moreover, per-pixel classification has become the focus of research interests in this regard. However, the efficient and effective processing and the interpretation of VHR satellite images remain a critical task. Convolutional neural networks (CNNs) have recently been applied to VHR satellite images with considerable success. However, the prevalent CNN models accept input data of fixed sizes and train the classifier using features extracted directly from the convolutional stages or the fully connected layers, which cannot yield pixel-to-pixel classifications. Moreover, training a CNN model requires large amounts of labeled reference data. These are challenging to obtain because per-pixel labeled VHR satellite images are not open access. In this paper, we propose a framework called the unsupervised-restricted deconvolutional neural network (URDNN). It can solve these problems by learning an end-to-end and pixel-to-pixel classification and handling a VHR classification using a fully convolutional network and a small number of labeled pixels. In URDNN, supervised learning is always under the restriction of unsupervised learning, which serves to constrain and aid supervised training in learning more generalized and abstract feature. To some degree, it will try to reduce the problems of overfitting and undertraining, which arise from the scarcity of labeled training data, and to gain better classification results using fewer training samples. It improves the generality of the classification model. We tested the proposed URDNN on images from the Geoeye and Quickbird sensors and obtained satisfactory results with the highest overall accuracy (OA) achieved as 0.977 and 0.989, respectively. Experiments showed that the combined effects of additional kernels and stages may have produced better results, and two-stage URDNN consistently produced a more stable result. We compared URDNN with four methods and found that with a small ratio of selected labeled data items, it yielded the highest and most stable results, whereas the accuracy values of the other methods quickly decreased. For some categories with fewer training pixels, accuracy for categories from other methods was considerably worse than that in URDNN, with the largest difference reaching almost 10%. Hence, the proposed URDNN can successfully handle the VHR image classification using a small number of labeled pixels. Furthermore, it is more effective than state-of-the-art methods.

This paper focuses on the application of a semi-automatic unsupervised change detection algorithm called Cross Correlation Analysis (CCA) to the detection of (semi-) natural grasslands changes at Very High Resolution (VHR). A reference validated Land Cover/Land Use map at time T1 and only one satellite image at time T2, with T2>T1, are required to detect changes occurred at T2 in the selected target class. This approach offers the possibility to reduce the costs of change detection when the acquisition of multi-seasonal VHR images at time T2 for supervised change detection is too expensive or when no archive VHR image is available in the past for unsupervised comparison between T1 and T2 images. A summer Worldview-2 image for a Natura 2000 test site was considered and the results appear encouraging.

Satellite imagery has been used to improve scientific research worldwide. In this study, the southern right whale (Eubalaena australis) was chosen to test the use of medium, high, and very high resolution (VHR) satellite images, on the Brazilian breeding ground. These images were used to identify the whales and were compared to aerial survey data collected in the same area. The VHR satellite images from the Pleiades‐1A satellite, available on Google Earth, displayed the best results when compared to those from Sentinel 2, Landsat 8, Rapid Eye, and Planet Scope. No significant differences were observed (Mann‐Whitney U test) between the possible whales recorded in the satellite images and of real whales recorded in situ by aerial surveys, while considering either the number of groups (p = .841, n = 5) or the total number of animals (p = .222, n = 5). Further, when using VHR images, the geographical positions of the whales recorded in situ had a positive correlation with the positions generated from the satellite images (Mantel test: r = 0.52, p = .001, n = 13). This technique may represent an important tool for detecting right whales, especially in countries where research funding is scarce.

Urban Land Use/Land Cover (LULC) information is essential for urban and environmental management. It is, however, very difficult to automatically extract detailed urban LULC information from remote sensing imagery, especially for a large urban area. Medium resolution imagery, such as Landsat Thematic Mapper (TM) data, cannot uncover detailed LULC information. Further, very high resolution (VHR) satellite imagery, such as IKONOS and QuickBird data, can only be applied to a small area, largely due to the data unavailability and high computation cost. As a result, little research has been conducted to extract detailed urban LULC information for a large urban area. This study, therefore, developed a three-layer classification scheme for deriving detailedurban LULC information by integrating newly launched Chinese GF-1 (medium resolution) and GF-2 (very high resolution) satellite imagery and synthetically incorporating geometry, texture, and spectral information through multi-resolution image segmentation and object-based image classification (OBIA). Homogeneous urban LULC types such as water bodies or large areas of vegetation could be derived from GF-1 imagery with 16 m and 8 m spatial resolutions, while heterogeneous urban LULC types such as industrial buildings, residential buildings, and roads could be extracted from GF-2 imagery with 3.2 m and 0.8 m spatial resolutions. The multi-resolution segmentation method and a random forest algorithm were employed to perform image segmentation and object-based image classification, respectively. An analysis of the results suggests an overall accuracy of 0.89 and 0.87 were achieved for the second and third level urban LULC classification maps, respectively. Therefore, the three-layer classification scheme has the potential to derive high accuracy urban LULC information through integrating medium and high-resolution remote sensing imagery.

The habitat distribution of Eastern High Arctic-Baffin Bay (EHA) beluga whales Del phinapterus leucas remains poorly understood due to their large range and remote summering areas. Incorporating social grouping characteristics into habitat analyses can provide insights into how whales use different environments. This study aimed to assess the significance of summering habitats to EHA beluga whales based on observed social characteristics. We used very high resolution (VHR) satellite imagery from summer 2020 to model beluga whale habitat use and social group size, cohesion, and composition across 3 sites: Creswell Bay, southwestern Devon Island estuaries , and southeastern Prince of Wales Island estuaries (PWI). Generalized additive and linear models were used to identify habitat preferences. Beluga whales preferred shallow, nearshore waters in Creswell Bay and PWI, likely due to opportunities for skin moulting through contact with the benthos. In PWI, whales were also associated with large river inputs, suggesting use of productive freshwater-saltwater mixing zones. Lone whales in PWI were more often found in deeper waters compared to groups of whales, potentially accessing higher-quality prey. In Devon Island, large groups (>5 whales) were commonly found in shallow areas, corresponding with schooling of preferred prey distributions. Group composition could only be determined in Creswell Bay, where no habitat segregation was observed between adult and mixed-age groups. Our findings demonstrate the value of VHR satellite imagery for studying beluga whale habitat use in remote regions. Understanding how beluga whales use different summering areas is crucial for assessing potential impacts of future habitat change and informing management decisions.