As one of the most destructive natural disasters, earthquakes have struck many countries around the world in recent years, causing serious economic losses. Change detection (CD) can be applied to post-earthquake building CD as it can infer interested change regions from multi-temporal remote sensing images. Furthermore, the CD with short imaging interval will better satisfy the needs of the emergency rescues after earthquakes. However, the capability of current methods built on deep neural networks is limited because the dataset with short imaging interval is absent. To meet post-disaster immediate relief, we create a CD dataset, the Turkey earthquake CD dataset (TUE-CD), for the detection of building collapse in the short term after an earthquake. Due to the high requirement for timeliness of post-event images, the orbit of the satellite during post-event imaging deviates from that during pre-event imaging, which leads to a side-looking problem between bi-temporal images. To deal with these challenges, we present a multi-scale feature interaction network (MSI-Net) for efficient interaction between bi-temporal features, as well as mitigating the effect of side-looking problems. Specifically, the proposed MSI-Net consists of joint cross-attention (JCA) modules, multi-scale offset calibration (MOC) modules, and feature integration (FeI) modules. The JCA module unifies channel cross-attention and spatial joint attention for sufficient feature interaction. The MOC module further estimates the offsets to align the bi-temporal image with the multi-scale features. Finally, calibrated features and multi-scale features are fused by FeI modules for the prediction of changed areas. The best mF1 and mIoU scores are achieved on two public datasets and the constructed TUE-CD dataset: WHU-CD (95.58%, 91.81%), CLCD (82.96%, 73.53%), and TUE-CD (78.02%, 68.48%). Experimental results demonstrate that the proposed MSI-Net provides competitive performance compared to the state-of-the-art CD methods. The TUE-CD Dataset as well as the code of MSI-Net will be available at https://github.com/RSMagneto/MSI-Net.
This paper aims to effectively preserve the spatial and spectral information of hyperspectral (HS) images. For this purpose, the author proposed a new image fusion method based on coupled tensor decomposition (CTD). First, the wanted high spatial resolution HS (HRHS) images were decomposed into the core tensor and basis matrices. Assuming that the basis matrices can be calculated from low spatial resolution HS (LRHS) images, the core tensor was estimated from the high spatial resolution multispectral (HRMS) images based on the relationship between HRHS and HRMS images. Finally, the HRHS images were obtained by reconstructing the core tensor with basis matrices. Owing to the good properties of tensor, the proposed method achieved better fusion results on different data sets than those of the contrastive methods. The research findings shed new light on hyperspectral pansharpening.
The perception layer of Internet of Things (IOT) consists of various sensors. It is the source of the IOT to identify objects and collect information. Information fusion collected from multi-sensor has been widely used in various fields, such as intelligent industry, intelligent agriculture, intelligent transportation, and intelligent environmental protection. In this paper, multi-sensor image fusion, multispectral (MS) and panchromatic (PAN) images, is studied, and the fused images are used in target detection, recognition, and classification. However, traditional methods based on an injection model generally consider the MS images as a whole to compute the spectral weights. They ignore the local information of MS images and produce some spectral distortions, because for different objects, the spectral response will be different. Therefore, we propose a novel multi-sensor image fusion based on application layer of IOT (IFIOT) to preserve the spectral information of MS images. In this method, local homogeneous areas are found first by superpixel segmentation. Due to good properties of superpixel, the homogeneous areas are uniform and contain only one kind of object. Then, we estimate the spectral weights for different bands on the homogeneous area. The injection gain has an important influence on fusion results. Therefore, we adaptively compute the gain coefficients by minimizing the error between the spectral degraded MS and PAN images. Finally, after the injection of spatial details obtaining from the PAN image, fused images are produced. Experimental results reveal that the IFIOT method can give good fusion results and the spectral information is preserved well.
In this letter, we proposed a multiscale integration network with quaternion convolution (MQ-Net) for the fusion of low spatial resolution multispectral (LRMS) and panchromatic (PAN) images. In this network, LRMS and PAN images are resampled at different scales and fed into feature fusion modules (FFMs) to merge the spatial and spectral information among them. Then, multiscale feature enhancement modules (MFEMs) are designed to sufficiently learn the spatial and spectral information at different scales. Meanwhile, we employ a quaternion convolution module (QCM) to better capture the dependencies within spectral bands of LRMS images. Then, the quaternion features are introduced into MFEMs for efficient feature enhancement. Finally, all information from different scales is integrated for the reconstruction of high LRMS images. Reduced- and full-resolution experiments are performed on GeoEye-1 and WorldView-2 satellite datasets. Compared to some state-of-the-art pansharpening methods, the proposed MQ-Net obtains better results in terms of qualitative and quantitative evaluations. The code is available at https://github.com/RSMagneto/MQ-Net .
13 August 2018 in Multimedia Tools and Applications