Adjacent Features Research Articles

Photovoltaic Image Registration Based on Feature Matching via Guided Spatial Consensus

Due to the high similarity of photovoltaic images, the difficulty of registering photovoltaic images increases while the accuracy requirement for registering images is high. This article proposed a guided spatial consensus registration algorithm based on local structural similarity constraint. First, local similarity structural constraints are established based on the characteristics of the neighborhood structure among adjacent matching feature points. Second, outliers are eliminated based on the proposed similarity structural constraints to obtain a high-precision matching feature set among a strictly constrained initial matching feature set. Third, the corresponding geometric parameter of the high-precision matching feature set is used as guided information for the selection of the final inlier set from a weakly constrained initial matching feature set. Finally, the proposed method is validated by photovoltaic images with brightness and geometric differences. The experimental results demonstrate the robustness and efficiency of the proposed algorithm in the application of photovoltaic images.

Hierarchical crack buffering triples ductility in eutectic herringbone high-entropy alloys.

In human-made malleable materials, microdamage such as cracking usually limits material lifetime. Some biological composites, such as bone, have hierarchical microstructures that tolerate cracks but cannot withstand high elongation. We demonstrate a directionally solidified eutectic high-entropy alloy (EHEA) that successfully reconciles crack tolerance and high elongation. The solidified alloy has a hierarchically organized herringbone structure that enables bionic-inspired hierarchical crack buffering. This effect guides stable, persistent crystallographic nucleation and growth of multiple microcracks in abundant poor-deformability microstructures. Hierarchical buffering by adjacent dynamic strain-hardened features helps the cracks to avoid catastrophic growth and percolation. Our self-buffering herringbone material yields an ultrahigh uniform tensile elongation (~50%), three times that of conventional nonbuffering EHEAs, without sacrificing strength.

Machine learning predictions of critical heat fluxes for pillar-modified surfaces

For convection research, one important topic is the maximum heat flux achieved on a boiling surface, known as the critical heat flux (CHF). This phenomenon is characterized by the formation of a blanket of heat-blocking vapor on the surface. Over several decades, numerous surface structures have been fabricated to enhance the CHF for various high-power cooling applications. However, the complexity of the surface structures and many other factors (e.g., capillary wicking flux) restrict the prediction of the CHF using theoretical models. In this work, three popular machine learning (ML) methods are employed to analyze and further predict the CHFs for a given surface modified with micro-structures. Among these, the random forest regression method consistently produced the best fitting models of previously published data. The importance analysis algorithm developed for random forest models facilitated efficient discovery of the most important descriptors predicting the CHF. One key descriptor used in these models was the mean beam length (MBL), a terminology borrowed from radiative heat transfer, which effectively described the characteristic spacing between adjacent surface features. The models showed greatest sensitivity to the MBL and the height of the features, compared to the other surface descriptors.

Hyperspectral image spectral-spatial classification via weighted Laplacian smoothing constraint-based sparse representation.

As a powerful tool in hyperspectral image (HSI) classification, sparse representation has gained much attention in recent years owing to its detailed representation of features. In particular, the results of the joint use of spatial and spectral information has been widely applied to HSI classification. However, dealing with the spatial relationship between pixels is a nontrivial task. This paper proposes a new spatial-spectral combined classification method that considers the boundaries of adjacent features in the HSI. Based on the proposed method, a smoothing-constraint Laplacian vector is constructed, which consists of the interest pixel and its four nearest neighbors through their weighting factor. Then, a novel large-block sparse dictionary is developed for simultaneous orthogonal matching pursuit. Our proposed method can obtain a better accuracy of HSI classification on three real HSI datasets than the existing spectral-spatial HSI classifiers. Finally, the experimental results are presented to verify the effectiveness and superiority of the proposed method.

Multi-Scale Feature Aggregation by Cross-Scale Pixel-to-Region Relation Operation for Semantic Segmentation

Exploiting multi-scale features has shown great potential in tackling semantic segmentation problems. The aggregation is commonly done with sum or concatenation (concat) followed by convolutional (conv) layers. However, it fully passes down the high-level context to the following hierarchy without considering their interrelation. In this work, we aim to enable the low-level feature to aggregate the complementary context from adjacent high-level feature maps by a cross-scale pixel-to-region relation operation. We leverage cross-scale context propagation to make the long-range dependency capturable even by the high-resolution low-level features. To this end, we employ an efficient feature pyramid network to obtain multi-scale features. We propose a Relational Semantics Extractor (RSE) and Relational Semantics Propagator (RSP) for context extraction and propagation respectively. Then we stack several RSP into an RSP head to achieve the progressive top-down distribution of the context. Experiment results on two challenging datasets Cityscapes and COCO demonstrate that the RSP head performs competitively on both semantic segmentation and panoptic segmentation with high efficiency. It outperforms DeeplabV3 <xref ref-type="bibr" rid="ref1">[1]</xref> by 0.7&#x0025; with 75&#x0025; fewer FLOPs (multiply-adds) in the semantic segmentation task.

Design of millimeter-wave chipless tag based on variable angle arc resonator

Faced with the development of miniaturization and compactness of chipless radio frequency identification (RFID) tags, and to reduce the cost of traditional chip tags, this paper proposes a miniaturized, polarization-insensitive, passive, and chipless based on a variable-angle arc resonator. Chipless tags have a high coding density of 44.44 bits/cm2 in the frequency range of 20GHz∼31GHz, and a coding capacity of 4.04 bits/cm2/GHz. The patch on the substrate determines the signature of each resonant frequency, thereby determining the number of identification data bits. To accommodate more data bits and pass barriers between resonant features, the metal patches are orthogonally polarized so that adjacent resonant features will not interfere with each other. As the mm-Waves ultra-wideband (UWB) chipless tags, it can be easily printed on various IoT applications and has great potential for low cost and identity verification. This work makes it easier to implement highly robust, inexpensive, and flexible chipless tags for IoT applications in the mm-Wave band.

Pyramidal Feature Shrinking for Salient Object Detection

Recently, we have witnessed the great progress of salient object detection (SOD), which benefits from the effectiveness of various feature aggregation strategies. However, existing methods usually aggregate the low-level features containing details and the high-level features containing semantics over a large span, which introduces noise into the aggregated features and generate inaccurate saliency map. To address this issue, we propose pyramidal feature shrinking network (PFSNet), which aims to aggregate adjacent feature nodes in pairs with layer-by-layer shrinkage, so that the aggregated features fuse effective details and semantics together and discard interference information. Specifically, pyramidal shrinking decoder (PSD) is proposed to aggregate adjacent features hierarchically in an asymptotic manner. Unlike other methods that aggregate features with significantly different information, this method only focuses on adjacent feature nodes in each layer and shrinks them to a final unique feature node. Besides, we propose adjacent fusion module (AFM) to perform mutual spatial enhancement between the adjacent features so as to dynamically weight the features and adaptively fuse the appropriate information. In addition, scale-aware enrichment module (SEM) based on the features extracted from backbone is utilized to obtain rich scale information and generate diverse initial features with dilated convolutions. Extensive quantitative and qualitative experiments demonstrate that the proposed intuitive framework outperforms 14 state-of-the-art approaches on 5 public datasets.

Learning Comprehensive Motion Representation for Action Recognition

For action recognition learning, 2D CNN-based methods are efficient but may yield redundant features due to applying the same 2D convolution kernel to each frame. Recent efforts attempt to capture motion information by establishing inter-frame connections while still suffering the limited temporal receptive field or high latency. Moreover, the feature enhancement is often only performed by channel or space dimension in action recognition. To address these issues, we first devise a Channel-wise Motion Enhancement (CME) module to adaptively emphasize the channels related to dynamic information with a channel-wise gate vector. The channel gates generated by CME incorporate the information from all the other frames in the video. We further propose a Spatial-wise Motion Enhancement (SME) module to focus on the regions with the critical target in motion, according to the point-to-point similarity between adjacent feature maps. The intuition is that the change of background is typically slower than the motion area. Both CME and SME have clear physical meaning in capturing action clues. By integrating the two modules into the off-the-shelf 2D network, we finally obtain a Comprehensive Motion Representation (CMR) learning method for action recognition, which achieves competitive performance on Something-Something V1 &amp; V2 and Kinetics-400. On the temporal reasoning datasets Something-Something V1 and V2, our method outperforms the current state-of-the-art by 2.3% and 1.9% when using 16 frames as input, respectively.

Lifting Scheme-Based Sparse Density Feature Extraction for Remote Sensing Target Detection

The design of backbones is of great significance for enhancing the location and classification precision in the remote sensing target detection task. Recently, various approaches have been proposed on altering the feature extraction density in the backbones to enlarge the receptive field, make features prominent, and reduce computational complexity, such as dilated convolution and deformable convolution. Among them, one of the most widely used methods is strided convolution, but it loses the information about adjacent feature points which leads to the omission of some useful features and the decrease of detection precision. This paper proposes a novel sparse density feature extraction method based on the relationship between the lifting scheme and convolution, which improves the detection precision while keeping the computational complexity almost the same as the strided convolution. Experimental results on remote sensing target detection indicate that our proposed method improves both detection performance and network efficiency.

Learning U-Net Based Multi-Scale Features in Encoding-Decoding for MR Image Brain Tissue Segmentation.

Accurate brain tissue segmentation of MRI is vital to diagnosis aiding, treatment planning, and neurologic condition monitoring. As an excellent convolutional neural network (CNN), U-Net is widely used in MR image segmentation as it usually generates high-precision features. However, the performance of U-Net is considerably restricted due to the variable shapes of the segmented targets in MRI and the information loss of down-sampling and up-sampling operations. Therefore, we propose a novel network by introducing spatial and channel dimensions-based multi-scale feature information extractors into its encoding-decoding framework, which is helpful in extracting rich multi-scale features while highlighting the details of higher-level features in the encoding part, and recovering the corresponding localization to a higher resolution layer in the decoding part. Concretely, we propose two information extractors, multi-branch pooling, called MP, in the encoding part, and multi-branch dense prediction, called MDP, in the decoding part, to extract multi-scale features. Additionally, we designed a new multi-branch output structure with MDP in the decoding part to form more accurate edge-preserving predicting maps by integrating the dense adjacent prediction features at different scales. Finally, the proposed method is tested on datasets MRbrainS13, IBSR18, and ISeg2017. We find that the proposed network performs higher accuracy in segmenting MRI brain tissues and it is better than the leading method of 2018 at the segmentation of GM and CSF. Therefore, it can be a useful tool for diagnostic applications, such as brain MRI segmentation and diagnosing.

Context-Aware Attentional Graph U-Net for Hyperspectral Image Classification

Hyperspectral image (HSI) registers hundreds of spectral bands, whose intraclass variability and interclass similarity are resourceful information to be mined. Intraclass variability reflects the nonuniform and redundancy of the spatial and semantic features extracted from HSI. Interclass similarity represents the inherent relationship between adjacent features and snapshots. Existing models extract the superficial correlation representation for HSI to tackle the classification task but fail to embed the interclass and intraclass correlations due to these models' intrinsic bottlenecks. Confronting the challenges of capturing interrelation for complex data in practice, we propose a Context-Aware Attentional Graph U-Net (CAGU) to improve these two modes of representation, which is more flexible in feature enhancement. In this method, attentional Graph U-Net is capable of extracting the intraclass embeddings within a non-Euclidean space by combining similar distributing feature vertices. The gated recurrent unit (GRU) is another critical component of our model to capture the context-aware dynamic interclass embeddings. Extensive experiments demonstrate that our model can efficiently outperform state-of-the-art methods across-the-board on five wide-adopted public data sets, namely, Pavia University, Indian Pines, Salinas Scene-show, Houston 2013, and Houston 2018, on par with the same scale of model parameters.

Dual-Path Sparse Hierarchical Network for Semantic Segmentation of Remote Sensing Images

Semantic segmentation of remote sensing images aims to label every pixel with the correct semantic category. The core challenge of the current deep convolutional network (ConvNet)-based methods lies in the difficulty of effectively aggregating high-level categorical semantics and low-level local details along the hierarchy of backbone. Most current approaches consider only fusing adjacent feature layers gradually with short-range feature connections, which lack the diversity of feature interactions, such as long-range cross-scale connections. To this end, we propose a novel dual-path sparse hierarchical network that is characterized by rich cross-scale feature interactions. Multiscale features are first sparsely grouped with a predefined interval, which is then aggregated via both long-range and short-range cross-scale connections in a hierarchical manner. Moreover, in order to further enrich the diversity of feature interactions, we also introduce another fusion path in parallel but with different sparsity for feature grouping, forming a dual-path network. In this way, our model is able to effectively aggregate multilevel features by incorporating both long-range and short-range feature interactions in both parallel and hierarchical manner. Meanwhile, the semantic and resolution gap between multilevel features can also be bridged.

Lip Reading using Local-Adjacent Feature Extractor and Multi-Level Feature Fusion

Lip reading is a widely used technology, which aims to infer text content from visual information. To represent lip information more efficiently and reduce the network parameters, most networks will first extract features from lip images and then classify the features. In recent studies, most researchers adopt convolutional networks to extract information from pixels which contain a lot of useless information, limiting the improvement of model accuracy. In this paper, we designed a graph structures and a lip segmentation network to effectively represent changes in the shape of the lips in adjacent frames and the ROI in local frame and propose two feature extractors, named U-net-based local feature extractor and graph-based adjacent feature extractor. We proposed a very challenging dataset to simulate extreme environments, including highly variable face properties, light intensity and so on. Finally, we designed several different levels of feature fusion methods. The experimental results on the proposed challenging dataset show that the model can effectively extract the useful information from content irrelevant information very well. The accuracy of our proposed model is 9.1% higher than that of baseline. This result shows that our proposed model can better adapt to the application of the wild environment.

An effective biometric identification system using enhanced palm texture features

Pattern recognition technologies aim to develop new techniques for transforming original data into other representations that honestly illustrate their content and can then be used to analyze and understand the original data. In the field of computer vision based on images, which is the dominant technology for machine vision solutions, image texture is a rich information source, used in many feature extraction techniques to obtain salient and distinctive features that were later used in pattern recognition applications. However, depending on the application, the extracted feature vector may be subject to some constraints such as length and accuracy. This work aims to develop a new texture-based feature extraction technique that can provide a feature vector with a high degree of distinction that can be controlled in terms of length and accuracy. The new scheme, called adjustable local binary pattern (A-LBP) or adjacent block features based LBP, derived from the original LBP, uses neighboring blocks and a linear relationship between the features of these blocks to form a binary codeword’s that can be used to represent the image. In our study, the block feature is extracted using the discrete cosine transform. To evaluate the performance of the proposed scheme, an A-LBP based biometric and bio-watermarking systems were developed.

Assessing Effects of Near-Field Synergistic Light Absorption on Ordered Inorganic Phototropic Growth.

We report herein that synergistic light absorption in the optical near-field enables nanoscale self-organization during inorganic phototropic growth. Se-Te was grown electrochemically under illumination from an incoherent, unstructured light source in geometrically constrained, wavelength scale areas. Despite the limited dimensions, with as few as two discrete features produced in a single sub-micron dimension, the deposit morphology exhibited defined order and anisotropy. Computer modeling analysis of light absorption in simulated structures revealed a synergy wherein light capture in a nanoscale feature was enhanced by the presence of additional adjacent features, with the synergistic effect originating predominantly from nearest neighbor contributions. Modeling moreover indicated that synergistic absorption is produced by scattering of the incident illumination by individual nanoscale features, leading to a local increase in the near-field intensity and consequently increasing the absorption in neighboring features. The interplay between these optical processes establishes the basis for spontaneous order generation via inorganic phototropic growth.

SCS-Net: A Scale and Context Sensitive Network for Retinal Vessel Segmentation.

Accurately segmenting retinal vessel from retinal images is essential for the detection and diagnosis of many eye diseases. However, it remains a challenging task due to (1) the large variations of scale in the retinal vessels and (2) the complicated anatomical context of retinal vessels, including complex vasculature and morphology, the low contrast between some vessels and the background, and the existence of exudates and hemorrhage. It is difficult for a model to capture representative and distinguishing features for retinal vessels under such large scale and semantics variations. Limited training data also make this task even harder. In order to comprehensively tackle these challenges, we propose a novel scale and context sensitive network (a.k.a., SCS-Net) for retinal vessel segmentation. We first propose a scale-aware feature aggregation (SFA) module, aiming at dynamically adjusting the receptive fields to effectively extract multi-scale features. Then, an adaptive feature fusion (AFF) module is designed to guide efficient fusion between adjacent hierarchical features to capture more semantic information. Finally, a multi-level semantic supervision (MSS) module is employed to learn more distinctive semantic representation for refining the vessel maps. We conduct extensive experiments on the six mainstream retinal image databases (DRIVE, CHASEDB1, STARE, IOSTAR, HRF, and LES-AV). The experimental results demonstrate the effectiveness of the proposed SCS-Net, which is capable of achieving better segmentation performance than other state-of-the-art approaches, especially for the challenging cases with large scale variations and complex context environments.

An Algorithm for Detecting the Integrity of Outer Frame Protection Net on Construction Site Based on Improved SSD

In recent years, the continuous development of the basic construction industry has promoted the progress of urban social economy, and the safety issues in construction have attracted more and more attention. The integrity of the external protection net is of great significance to the safety of the construction personnel. Most of the traditional methods of detecting damage to the protection net have problems such as poor real-time performance and low detection accuracy. In this paper, we present a novel approach based on SSD for solving the problem of the integrity detection of the external frame protection net in the construction site. The improved network merges adjacent feature layers to enrich the semantics of shallow feature maps, thereby improving the detection accuracy of the network, and simultaneously introduces a shallower feature layer and a more advanced receptive field module to improve the detection ability of small targets. The experimental results show that compared with the SSD algorithm, the improved algorithm has higher detection accuracy in the detection task of the external protective net on the construction site, and can meet the real-time detection requirements.

Attention Modulated Multiple Object Tracking with Motion Enhancement and Dual Correlation

The one-shot multiple object tracking (MOT) framework has drawn more and more attention in the MOT research community due to its advantage in inference speed. However, the tracking accuracy of current one-shot approaches could lead to an inferior performance compared with their two-stage counterparts. The reasons are two-fold: one is that motion information is often neglected due to the single-image input. The other is that detection and re-identification (ReID) are two different tasks with different focuses. Joining detection and re-identification at the training stage could lead to a suboptimal performance. To alleviate the above limitations, we propose a one-shot network named Motion and Correlation-Multiple Object Tracking (MAC-MOT). MAC-MOT introduces a motion enhance attention module (MEA) and a dual correlation attention module (DCA). MEA performs differences on adjacent feature maps which enhances the motion-related features while suppressing irrelevant information. The DCA module focuses on decoupling the detection task and re-identification task to strike a balance and reduce the competition between these two tasks. Moreover, symmetry is a core design idea in our proposed framework which is reflected in Siamese-based deep learning backbone networks, the input of dual stream images, as well as a dual correlation attention module. Our proposed approach is evaluated on the popular multiple object tracking benchmarks MOT16 and MOT17. We demonstrate that the proposed MAC-MOT can achieve a better performance than the baseline state of the arts (SOTAs).

Classification of Very-High-Spatial-Resolution Aerial Images Based on Multiscale Features with Limited Semantic Information

Recently, deep learning has become the most innovative trend for a variety of high-spatial-resolution remote sensing imaging applications. However, large-scale land cover classification via traditional convolutional neural networks (CNNs) with sliding windows is computationally expensive and produces coarse results. Additionally, although such supervised learning approaches have performed well, collecting and annotating datasets for every task are extremely laborious, especially for those fully supervised cases where the pixel-level ground-truth labels are dense. In this work, we propose a new object-oriented deep learning framework that leverages residual networks with different depths to learn adjacent feature representations by embedding a multibranch architecture in the deep learning pipeline. The idea is to exploit limited training data at different neighboring scales to make a tradeoff between weak semantics and strong feature representations for operational land cover mapping tasks. We draw from established geographic object-based image analysis (GEOBIA) as an auxiliary module to reduce the computational burden of spatial reasoning and optimize the classification boundaries. We evaluated the proposed approach on two subdecimeter-resolution datasets involving both urban and rural landscapes. It presented better classification accuracy (88.9%) compared to traditional object-based deep learning methods and achieves an excellent inference time (11.3 s/ha).

Mining the features of spatial adjacency relationships to improve the classification of high resolution remote sensing images based on complex network

Due to the complexity of spatial and structural patterns of remote sensing images, land cover classification is still an active, ongoing problem in the field of image classification. The limited spectrum and texture features can be used often lead to low accuracy in remote sensing image classifications. Spatial structure information, which is commonly regarded as very useful information for improving land cover classifications, characterizes a composite of several types of objects or land covers. However, it is rarely used because it is difficult to describe and calculate. Generally, images contain the relationships of spatial adjacency, spatial direction and spatial distance among multiple objects. This study selects spatial adjacency relationships as an example, mines the features of spatial adjacency, and uses to improve classification. Therefore, this paper (1) presents a modeling method of remote sensing images based on complex networks to express the spatial structure information and make it computable firstly; (2) gives network expression and calculation methods of spectrum, texture and object features in images; (3) discusses a discovery method of spatial adjacency feature based on network. Finally, this paper gives a strategy to improve the classification for high-resolution image by using the spatial adjacency features. Two kinds of data in coastal zones and city areas are used to verify the method. The results show that the proposed approach, which considers the features of spatial adjacencies, improves the classification accuracy by at least 4% compared with the approaches that do not consider spatial adjacencies. Comparing with the results of deep learning method, our method has higher overall accuracy when the numbers of samples are small.