Search

Remote sensing image (RSI) scene classification is a hot topic in the field of remote sensing and has garnered a lot of attention. The key issue in image classification is effectively understanding semantic content. Convolutional neural networks (CNNs) are generally recognized to significantly improve classification performance due to their powerful feature extraction capabilities. However, the overall structure of the model is complicated and has a large number of parameters, making it difficult to extract more efficient features. To address these problems, in this paper, we propose a lightweight skip-connected expansion Inception network called SEINet. To capture characteristics at a more granular level, we create a new lightweight backbone network with fewer parameters based on the existing network architecture. Additionally, the paper introduces a skip-connected expansion Inception (SEI) module for extracting context-dependent relationships. The ablation experiments verify the effectiveness of our proposed module. Experiment findings on two public datasets demonstrate that our method has advantages in classification accuracy and execution efficiency over state-of-the-art (SOTA) methods. [ABSTRACT FROM AUTHOR]

Scene classification of high-resolution remote sensing (HRRS) images has drawn widespread attention as a challenging task. Convolutional neural networks(CNNs) have been widely applied in remote sensing classification. However, owing to tiny local objects and complex spatial distributions in HRRS images, existing vanilla CNN-based methods can rarely perceive the critical regions. In this paper, we present a novel context-aware and channel re-weighting network (CACRWNet) to improve classification performance by strengthening multilayer spatial semantics and achieving a proper balance between local and global semantics. First, the proposed framework utilizes a backbone with a noise suppression (NS) module to extract enhanced features. Then, a deep context-aware (DCA) module is constructed to exploit distinctive features and corresponding context information to enhance feature differentiation while reducing the loss of local details. Third, a novel channel re-weighting (CRW) module is designed to reallocate the weight of each channel whilst utilizing a fusion of high-level and low-level features. Experiments on two open remote sensing scene classification (RSSC) datasets (UCM and AID) show that our method achieves remarkable classification performance compared with some state-of-art algorithms. [ABSTRACT FROM AUTHOR]

Chi-squared transform (CST)-based methods are simple and effective methods for detecting changes in remotely sensed images that have been registered and aligned. The methods operate directly on information stored in the difference image. However, the estimated mean and covariance matrix of the Gaussian distribution that describes theunchangedpixels can be biased when thechangedpixels (outliers) are also included. To overcome this issue, we propose a pixel-based unsupervised change detection method that gives robust estimates of these parameters. The method is iterative but requires only a small number of iterations. In addition, we also design an algorithm to automatically search for the optimal threshold that is needed for classifying changedversusunchanged pixels. This algorithm finds the optimal threshold where the mean and covariance matrix of the change detection result most agree with those statistics obtained from the above-mentioned robust algorithm. We refer to our change detection method as the robust CST (RCST) method. The proposed method has been evaluated on two image data-sets and compared with four state-of-the-art methods. The effectiveness of RCST is confirmed by its low overall errors (OE) and high kappa coefficients on both data-sets. [ABSTRACT FROM PUBLISHER]

Semantic feature recognition in colour images is required for identifying uneven patterns in object detection and classification. The semantic features are identified by segmenting the colorimetric sensor array features through machine learning paradigms. Semantic segmentation is a method for identifying distinct elements in an image. This can be considered a task involving image classification at the pixel level. This article introduces a semantic feature-dependent array segmentation method (SFASM) to improve recognition accuracy due to irregular semantics. The proposed method incorporates a deep convolutional neural network for detecting the semantic and un-semantic features based on sensor array representations. The colour distributions per array are identified for horizontal and vertical semantics analysis. In this analysis, deep learning classifies the uneven patterns based on colour distribution, i.e. the consecutive and scattered colour distribution pixels in an array are correlated for their similarity. This similarity identification is maximized through max-pooling and recurrent iterations, preventing detection errors. The proposed method classifies the semantic features for further correlation sections, improving the accuracy. The proposed method's performance is thus validated using the metrics precision, analysis time and F1-Score. [ABSTRACT FROM AUTHOR]

Abstract: The classification problem of remotely sensed images with hyperspectral and hyperspatial resolution images is being paid more and more attention. The success of remotely sensed images classification depends on many facts, such as the availability of high-quality images and ancillary data, proper classification procedure, and the analytical ability of scientific researcher. Therefore, lots of methods of combing spatial, spectral and texture information were proposed. However, these methods may ignore these facts as below. On the one hand, many details of the original remotely sensed images may be covered up by the too much overlapping information. On the other hand, the classification process is time-consuming. Therefore, a new and efficient classification of remotely sensed images method is introduced to overcome these shortcomings. The proposed method deals with the original information provided by the remotely sensed images is considered. The block matrix is made of training samples of the same class. The details of original remotely sensed images is obtained from the QR decomposition with column pivoting (QRcp) or singular value decomposition (SVD). And then, using fisher linear discriminant analysis (FLDA) methods, the projection data information of original remotely sensed images is jointly used for the classification through a support vector machines (SVMs) formulation. Experiments on hyperspatial and hyperspectral images are performed to test and evaluate the effectiveness of the proposed method. [Copyright &y& Elsevier]

• A novel joint model enhances the semantic segmentation in remote sensing images. • An improved UNet has the ability to quickly converge and avoid neuron death. • A modified SegNet optimizes the segmentation for images with imprecise labels. • Real-world experiments demonstrate the effectiveness. Semantic segmentation for high-resolution (HR) remote sensing (RS) images under the condition of a small number of training samples with imprecise labels is a challenging task. In this paper, we propose a novel method based on improved UNet and SegNet. First, a batch normalization layer is introduced into the original UNet to accelerate the convergence speed and an ELU activation function is selected instead of ReLU to avoid the neuron death. Second, to enhance the conventional SegNet, the encoder is reconstructed and a skip connection is designed to reuse the deep features, which is also beneficial for the network convergence. Third, a joint model is constructed by combining the improved UNet and SegNet and meanwhile a voting strategy is proposed to compute the final results. Experiments on the real-world HR RS images verify the effectiveness and superiority of the proposed method. [Display omitted] [ABSTRACT FROM AUTHOR]

Single-video-based temporal super-resolution reconstruction (SRR) can get rid of constraints imposed on multiple-video-based temporal SRR, including device scarcity, temporal synchronization, complicated registration, and high cost. But frame interpolation in single-video-based temporal SRR cannot deblur well, and the selection or determination related to interpolation function has a lack of evidence, which reduces the fidelity. Additionally, the subsequent spatial deblurring is suboptimal for the motion blur in video because the formation of the motion blur is different from that of the spatial blur. This paper proposes a temporal SRR based on pixel stream and featured prior model to increase the frame rate and the definition of motion blurred single video. The proposed temporal SRR views the single video as a bundle of pixel streams and implements maximum-a-posteriori-based pixel stream SRR stream-by-stream without the selection or determination related to interpolation function. What is more, a featured temporal prior about the pixel stream is proposed and introduced into the pixel stream SRR, which can well remove the motion blur. Experimental results show the proposed temporal SRR is effective. ©2016 SPIE and IS&T [DOI: 10.1117/1.JEI.25.5.053011] [ABSTRACT FROM AUTHOR]