Extracting buildings from very high resolution (VHR) images has attracted much attention but is still challenging due to their large varieties in appearance and scale. Convolutional neural networks (CNNs) have shown effective superior performance automatically learning high-level discriminative features extracting buildings. However, the fixed receptive fields make conventional CNNs insufficient tolerate scale changes. Multiscale CNN (MCNN) a promising structure meet this challenge....

Physical unclonable function (PUF) is an advanced hardware security technology. Most conventional encryption approaches rely on the secure keys stored in nonvolatile memory, which are vulnerable to physical attacks. In contrast, PUF exploits fabrication variations generate keys. As there significant induced carbon nanotube (CNT)-based circuits, they natural candidates for building highly PUFs. However, existing PUFs reported be machine learning modeling this paper, we develop a novel CNT...

Coastal wetland mapping plays an essential role in monitoring climate change, the hydrological cycle, and water resources. In this study, a novel classification framework based on gravitational optimized multilayer perceptron classifier extended multi-attribute profiles (EMAPs) is presented for coastal using Sentinel-2 multispectral instrument (MSI) imagery. proposed method, morphological attribute (APs) are firstly extracted four filters characteristics of wetlands each band from These APs...

Classifying land use from postearthquake very high-resolution (VHR) images is challenging due to the complexity of objects in Earth surface after an earthquake. Convolutional neural network (CNN) exhibits satisfied performance differentiating complex objects, thanks its automatic extraction high-level features and accurate identification target geo-objects. Nevertheless, view scale variance natural fact that CNN suffers fixed receptive field, reduced feature resolution, insufficient training...

Band selection is an important data dimensionality reduction tool in hyperspectral images (HSIs). To identify the most informative subset band from hundreds of highly corrected bands HSIs, a novel method using crossover‐based gravitational search algorithm (CGSA) presented this study. In method, discriminative capability each evaluated by combined optimisation criterion, which constructed based on overall classification accuracy and size subset. As evolution updated V ‐shaped transfer...

Due to the intricate and diverse nature of industrial systems, traditional optimization algorithms require a significant amount time search for optimal solution throughout entire design space, making them unsuitable meeting practical demands. To address this issue, we propose novel approach that combines surrogate models with algorithms. Firstly, introduce Sparse Gaussian Process regression (SGP) into model, proposing SGP surrogate-assisted method. This effectively overcomes computational...

The high interior heterogeneity of land surface covers in high-resolution image coastal cities makes classification challenging. To meet this challenge, a Multi-Scale Superpixels-based Classification method using Optimized Spectral–Spatial features, denoted as OSS-MSSC, is proposed paper. In the method, multi-scale superpixels are firstly generated to capture local spatial structures ground objects with various sizes. Then, normalized difference vegetation index and extend multi-attribute...

GaoFen-2 (GF-2) is a new high resolution remote sensing satellite launched by China. The fine spatial of GF-2 makes it suitable for urban studies. However, the frequent occurrence shadows in images brings about great difficulties to practical applications. Therefore, accurate shadow detection necessary step before applications data. In this paper, we propose novel technique which combines spectral, and morphological attribute profiles (MAPs) features mapping building images. algorithm,...

The purpose of this study is mainly lies in using the RFID technology with function such identifying, tracking and repetition read-write characteristic, applies newborn newborn's identity identification, from delivery room birth until to discharged hospital. Through system establish, Improve accuracy drug delivery, guarantees identification effectively safety control.

Spectral clustering is a key research topic in the field of machine learning and data mining. Most existing spectral algorithms are built upon Gaussian Laplacian matrices, which sensitive to parameters. We propose novel parameter free, distance consistent Locally Linear Embedding. The proposed LLE promises that edges between closer points have greater weight.Furthermore, we improved via embedded label propagation. Our algorithm two advancements state art:1) propagation,which propagates nodeś...

Accurate detection of topographic shadows is great importance, since shadowing an inevitable hamper for the interpretation remotely sensed images covered mountainous areas. In this paper, a novel method proposed effective and efficient shadow obtained from Sentinel-2A multispectral imager (MSI) by combining both spectral spatial information. method, four feature indices were firstly extracted original bands to capture essential characteristics. Specifically, we constructed index (TSI)...

With the development of cloud computing, integrity data is becoming increasingly important. The auditing schemes for allow owners to verify stored in an untrusted server. Most public are based on key infrastructure (PKI), which may lead certificate management problems. Recently, identity-based scheme was proposed and it could effectively reduce computation cost auditors solve However, proved be insecure. In this paper, we consider malicious auditor propose a new against computing....

Extracting spectral-spatial information via sparse representation is a hotspot for hyperspectral image (HSI) classification. However, the extracted by traditional joint classification (JSRC) method affected heterogeneous and noisy pixels, which leads to some misclassifications. In this paper, we proposed framework based on superpixel-constrained weighted HSI Superpixel constraint firstly used remove effects of are located in fixed sized blocks adopted JSRC. The weighting scheme then...

Abstract Background Analysis of DNA microarray data usually begins with a normalization step where intensities different arrays are adjusted to the same scale so that intensity levels from can be compared one other. Both simple total array intensity-based as well more complex "local level" dependent methods have been developed, some which widely used. Much less developed for analysis include those bypass and therefore yield results not confounded by potential errors. Results Instead focusing...

Accurate estimation of impervious surfaces is important but challenging due to the spectral confusion between different land covers. Recently, synergistic use optical and Synthetic Aperture Radar (SAR) data has shown advantage in improving estimation. In this paper, multiple kernel learning (MKL) was employed combine heterogeneous features Landsat-8 Sentinel-1A data. Impervious surface estimated at a sub-pixel level based on support vector regression (SVR) model. The percentage (ISP)...

Lipid-binding proteinsjoin many important biological processes. proteins are highly related to diseases, such as metabolic cancer and autoimmune diseases. The existed studies of predictinglipid-binding functions or sites, but notidentify the lipid-binding ornot proteins.This study purpose a systematic approach identify small set physicochemical biochemical properties in AAindex database design support vector machine (SVM) based classifier for predicting analyzing proteins. merits this...