Search

Periodically monitoring the pavement conditions is of great importance to many intelligent transportation activities. Timely and correctly identifying the distresses or anomalies on pavement surfaces can help to smooth traffic flows and avoid potential threats to pavement securities. In this paper, we develop a novel context-augmented capsule feature pyramid network (CCapFPN) to detect cracks from pavement images. The CCapFPN adopts vectorial capsules to represent high-level, intrinsic, and salient features of cracks. By designing a feature pyramid architecture, the CCapFPN can fuse different levels and different scales of capsule features to provide a high-resolution, semantically strong feature representation for accurate crack detection. To take advantage of the context properties, a context-augmented module is embedded into each stage of the CCapFPN to rapidly enlarge the receptive field. The CCapFPN performs effectively and efficiently in processing pavement images of diverse conditions and detecting cracks of different topologies. Quantitative evaluations show that an overall performance with a precision, a recall, and an F-score of 0.9200, 0.9149, and 0.9174, respectively, were achieved on the test datasets. Comparative studies with some existing deep learning and edge based crack detection methods also confirm the superior performance of the CCapFPN in crack detection tasks.

The up-to-date and accurate building database serves as an important prerequisite to many applications. However, caused by the issues of shape and size variations, texture and distribution diversities, and occlusion and shadow covers of buildings in remote sensing images, it is still challenging to well guarantee the integrity and accuracy of the extracted building instances. This letter proposes a high-resolution capsule network (HR-CapsNet) to conduct building extraction. First, designed with an HR-CapsNet architecture assisted by multiresolution feature propagation and fusion, the HR-CapsNet can provide semantically strong and spatially accurate feature representations to promote the pixel-wise building extraction accuracy. In addition, integrated with an efficient capsule feature attention module, the HR-CapsNet can attend to channel-wise informative and class-specific spatial features to boost the feature encoding quality. Quantitative evaluations, visual inspections, and comparative experiments on two large remote sensing image datasets demonstrate that the HR-CapsNet provides a feasible and competitive solution to building extraction tasks.

Periodically conducting land cover mapping plays a vital role in monitoring the status and changes of the land use. The up-to-date and accurate land use database serves importantly for a wide range of applications. This letter constructs an efficient self-attention capsule network (ESA-CapsNet) for land cover classification of multispectral light detection and ranging (LiDAR) data. First, formulated with a novel capsule encoder-decoder architecture, the ESA-CapsNet performs promisingly in extracting high-level, informative, and strong feature semantics for pixel-wise land cover classification by using the five types of rasterized feature images. Furthermore, designed with a novel capsule-based attention module, the channel and spatial feature encodings are comprehensively exploited to boost the feature saliency and robustness. The ESA-CapsNet is evaluated on two multispectral LiDAR data sets and achieves an advantageous performance with the overall accuracy, average accuracy, and kappa coefficient of over 98.42%, 95.15%, and 0.9776, respectively. Comparative experiments with the existing methods also demonstrate the effectiveness and applicability of the ESA-CapsNet in land cover classification tasks.

The information-accurate road network database is greatly significant and provides essential input to many transportation-related activities. Recently, remote sensing images have been an important ...

Building footprint extraction plays an important role in a wide range of applications. However, due to size and shape diversities, occlusions, and complex scenarios, it is still challenging to accurately extract building footprints from aerial images. This letter proposes a capsule feature pyramid network (CapFPN) for building footprint extraction from aerial images. Taking advantage of the properties of capsules and fusing different levels of capsule features, the CapFPN can extract high-resolution, intrinsic, and semantically strong features, which perform effectively in improving the pixel-wise building footprint extraction accuracy. With the use of signed distance maps as ground truths, the CapFPN can extract solid building regions free of tiny holes. Quantitative evaluations on an aerial image data set show that a precision, recall, intersection-over-union (IoU), and F-score of 0.928, 0.914, 0.853, and 0.921, respectively, are obtained. Comparative studies with six existing methods confirm the superior performance of the CapFPN in accurately extracting building footprints.

Accurate road marking extraction and classification play a significant role in the development of autonomous vehicles (AVs) and high-definition (HD) maps. Due to point density and intensity variations from mobile laser scanning (MLS) systems, most of the existing thresholding-based extraction methods and rule-based classification methods cannot deliver high efficiency and remarkable robustness. To address this, we propose a capsule-based deep learning framework for road marking extraction and classification from massive and unordered MLS point clouds. This framework mainly contains three modules. Module I is first implemented to segment road surfaces from 3D MLS point clouds, followed by an inverse distance weighting (IDW) interpolation method for 2D georeferenced image generation. Then, in Module II, a U-shaped capsule-based network is constructed to extract road markings based on the convolutional and deconvolutional capsule operations. Finally, a hybrid capsule-based network is developed to classify different types of road markings by using a revised dynamic routing algorithm and large-margin Softmax loss function. A road marking dataset containing both 3D point clouds and manually labeled reference data is built from three types of road scenes, including urban roads, highways, and underground garages. The proposed networks were accordingly evaluated by estimating robustness and efficiency using this dataset. Quantitative evaluations indicate the proposed extraction method can deliver 94.11% in precision, 90.52% in recall, and 92.43% in F1-score, respectively, while the classification network achieves an average of 3.42% misclassification rate in different road scenes.

Object detection from remote sensing images serves as an important prerequisite to many applications. However, caused by scale and orientation variations, appearance and distribution diversities, o...

Although significant improvement has been achieved in fully autonomous driving and semantic high-definition map (HD) domains, most of the existing 3D point cloud segmentation methods cannot provide high representativeness and remarkable robustness. The principally increasing challenges remain in completely and efficiently extracting high-level 3D point cloud features, specifically in large-scale road environments. This paper provides an end-to-end feature extraction framework for 3D point cloud segmentation by using dynamic point-wise convolutional operations in multiple scales. Compared to existing point cloud segmentation methods that are commonly based on traditional convolutional neural networks (CNNs), our proposed method is less sensitive to data distribution and computational powers. This framework mainly includes four modules. Module I is first designed to construct a revised 3D point-wise convolutional operation. Then, a U-shaped downsampling-upsampling architecture is proposed to leverage both global and local features in multiple scales in Module II. Next, in Module III, high-level local edge features in 3D point neighborhoods are further extracted by using an adaptive graph convolutional neural network based on the K-Nearest Neighbor (KNN) algorithm. Finally, in Module IV, a conditional random field (CRF) algorithm is developed for postprocessing and segmentation result refinement. The proposed method was evaluated on three large-scale LiDAR point cloud datasets in both urban and indoor environments. The experimental results acquired by using different point cloud scenarios indicate our method can achieve state-of-the-art semantic segmentation performance in feature representativeness, segmentation accuracy, and technical robustness.

Vehicle logo recognition provides an important supplement to vehicle make and model analysis. Some of the existing vehicle logo recognition methods depend on the detection of license plates to roughly locate vehicle logo regions using prior knowledge. The vehicle logo recognition performance is greatly affected by the license plate detection techniques. This paper presents a cascaded deep convolutional network for directly recognizing vehicle logos without depending on the existence of license plates. This is a two-stage processing framework composed of a region proposal network and a convolutional capsule network. First, potential region proposals that might contain vehicle logos are generated by the region proposal network. Then, the convolutional capsule network classifies these region proposals into the background and different types of vehicle logos. We have evaluated the proposed framework on a large test set towards vehicle logo recognition. Quantitative evaluations show that a detection rate, a recognition rate, and an overall performance of 0.987, 0.994, and 0.981, respectively, are achieved. Comparative studies with the Faster R-CNN and other three existing methods also confirm that the proposed method performs effectively and robustly in recognizing vehicle logos of various conditions.

Timely and accurately measuring surface water bodies and monitoring their conditions and changes are greatly important to a wide range of environmental and social activities. Recently, with the dev...

Multispectral LiDAR (Light Detection And Ranging) is characterized of the completeness and consistency of its spectrum and spatial geometric data, which provides a new data source for land-cover classification. In recent years, the convolutional neural network (CNN), compared with traditional machine learning methods, has made a series of breakthroughs in image classification, object detection, and image semantic segmentation due to its stronger feature learning and feature expression abilities. However, traditional CNN models suffer from some issues, such as a large number of layers, leading to higher computational cost. To address this problem, we propose a CNN-based multi-spectral LiDAR land-cover classification framework and analyze its optimal parameters to improve classification accuracy. This framework starts with the preprocessing of multi-spectral 3D LiDAR data into 2D images. Next, a CNN model is constructed with seven fundamental functional layers, and its hyper-parameters are comprehensively discussed and optimized. The constructed CNN model with the optimized hyper-parameters was tested on the Titan multi-spectral LiDAR data, which include three wavelengths of 532 nm, 1064 nm, and 1550 nm. Extensive experiments demonstrated that the constructed CNN with the optimized hyper-parameters is feasible for multi-spectral LiDAR land-cover classification tasks. Compared with the classical CNN models (i.e., AlexNet, VGG16 and ResNet50) and our previous studies, our constructed CNN model with the optimized hyper-parameters is superior in computational performance and classification accuracies.

Land cover mapping is an effective way to quantify land resources and monitor their changes. It plays an important role in a wide range of applications. This letter proposes a hybrid capsule network for land cover classification using multispectral light detection and ranging (LiDAR) data. First, the multispectral LiDAR data were rasterized into a set of feature images to exploit the geometrical and spectral properties of different types of land covers. Then, a hybrid capsule network composed of an encoder network and a decoder network is trained to extract both high-level local and global entity-oriented capsule features for accurate land cover classification. Quantitative classification evaluations on two data sets show that the overall accuracy, average accuracy, and kappa coefficient of over 97.89%, 94.54%, and 0.9713, respectively, are obtained. Comparative studies with five existing methods confirm that the proposed method performs robustly and accurately in land cover classification using the multispectral LiDAR data.

Traffic-sign recognition plays an important role in road transportation systems. This letter presents a novel two-stage method for detecting and recognizing traffic signs from mobile Light Detection and Ranging (LiDAR) point clouds and digital images. First, traffic signs are detected from mobile LiDAR point cloud data according to their geometrical and spectral properties, which have been fully studied in our previous work. Afterward, the traffic-sign patches are obtained by projecting the detected points onto the registered digital images. To improve the performance of traffic-sign recognition, we apply a convolutional capsule network to the traffic-sign patches to classify them into different types. We have evaluated the proposed framework on data sets acquired by a RIEGL VMX-450 system. Quantitative evaluations show that a recognition rate of 0.957 is achieved. Comparative studies with the convolutional neural network (CNN) and our previous supervised Gaussian–Bernoulli deep Boltzmann machine (GB-DBM) classifier also confirm that the proposed method performs effectively and robustly in recognizing traffic signs of various types and conditions.

Effective object extraction plays an important role in many point cloud-based applications. This letter proposes a 3-D feature matching framework for point cloud object extraction. To determine the optimal affine transformation parameters for each template feature point, a convex dissimilarity function and the locally affine-invariant geometric constraints are designed to construct the overall objective function. The 3-D feature matching framework is integrated into a point cloud object extraction workflow. Extraction results on six test data sets show that average completeness, correctness, quality, and $F_{1}$ -measure of 0.96, 0.97, 0.93, and 0.96, respectively, are obtained in extracting light poles, vehicles, and palm trees. Comparative studies also confirm that the proposed method performs effectively and robustly, and exhibits superior or compatible performance over the other compared methods.

Object detection from remote sensing imagery plays a significant role in a wide range of applications, including urban planning, intelligent transportation systems, ecology and environment analysis, etc. However, scale variations, orientation variations, illumination changes, and partial occlusions, as well as image qualities, bring great challenges for accurate geospatial object detection. In this paper, we propose an efficient orientation guided anchoring based geospatial object detection network based on convolutional neural networks. To handle objects of varying sizes, the feature extraction subnetwork extracts a pyramid of semantically strong features at different scales. Based on orientation guided anchoring, the anchor generation subnetwork generates a small set of high-quality, oriented anchors as object proposals. After orientation region of interest pooling, objects of interest are detected from the object proposals through the object detection subnetwork. The proposed method has been tested on a large geospatial object detection dataset. Quantitative evaluations show that an overall completeness, correctness, quality, and F1-measure of 0.9232, 0.9648, 0.8931, and 0.9435, respectively, are obtained. In addition, the proposed method achieves a processing speed of 8 images per second on a GPU on the cloud computing platform. Comparative studies with the existing object detection methods also demonstrate the advantageous detection accuracy and computational efficiency of our proposed method.

Periodical road manhole cover measurement is extremely important to ensure road safety and reduce traffic disasters. This letter proposes an effective method for delineating road manhole covers from mobile laser scanning point cloud data. To improve processing efficiency, first, road surface points are segmented and rasterized into georeferenced intensity images. Then, object-oriented patches are generated through superpixel segmentation and further fed to a convolutional capsule network classifier for manhole cover detection. Finally, manhole covers are accurately delineated through a marked point process of disks. Quantitative evaluations on three data sets show that an average completeness, correctness, quality, and ${F} _{1}$ -measure of 0.965, 0.961, 0.929, and 0.963, respectively, are obtained. Comparative studies with three existing methods confirm that the proposed method performs superiorly in delineating manhole covers of varying conditions and on complex road surface environments.

Geometric feature acts as an important role in point cloud shape classification tasks. Previous methods have proved that the geometric information of point clouds effectively improves the classification accuracy. Mo-Net firstly introduced geometric moments into point cloud shape classification, which, to fit the form of second order geometric moments, extends the number of input channels from three to nine. Unfortunately, similar to PointNet, Mo-Net cannot capture the local structures. To address this issue, we propose a graph geometric moments convolution neural network (GGM-Net), which learns local geometric features from geometric moments representation of a local point set. The core module of the GGM-Net is to learn features from geometric moments (termed as GGM convolution). Specifically, the GGM convolution learns point features and local features from the first and second order geometric moments of points and its local neighbors, respectively, and then combines these features by using an addition operation. In this way, a geometrical local representation about points is obtained, which leads to much surface geometry awareness and robustness. Equipped with the GGM convolution, GGM-Net, a simple end-to-end architecture, is developed to achieve a competitive accuracy on the benchmark dataset ModelNet40 and perform more efficiently in terms of memory and computational complexity.

Vehicle detection plays an important role in a variety of traffic-related applications. However, due to the scale and orientation variations and partial occlusions of vehicles, it is still challengeable to accurately detect vehicles from remote sensing images. This letter proposes a convolutional capsule network for detecting vehicles from high-resolution remote sensing images. First, a test image is segmented into superpixels to generate meaningful and nonredundant patches. Then, these patches are input to a convolutional capsule network to label them into vehicles or the background. Finally, nonmaximum suppression is adopted to eliminate repetitive detections. Quantitative evaluations on four test data sets show that average completeness, correctness, quality, and F1-measure of 0.93, 0.97, 0.90, and 0.95, respectively, are obtained. Comparative studies with three existing methods confirm that the proposed method effectively performs in detecting vehicles of various conditions.

A multispectral light detection and ranging (LiDAR) system, which simultaneously collects spatial geometric data and multi-wavelength intensity information, opens the door to three-dimensional (3-D) point cloud classification and object recognition. Because of the irregular distribution property of point clouds and the massive data volume, point cloud classification directly from multispectral LiDAR data is still challengeable and questionable. In this paper, a point-wise multispectral LiDAR point cloud classification architecture termed as SE-PointNet++ is proposed via integrating a Squeeze-and-Excitation (SE) block with an improved PointNet++ semantic segmentation network. PointNet++ extracts local features from unevenly sampled points and represents local geometrical relationships among the points through multi-scale grouping. The SE block is embedded into PointNet++ to strengthen important channels to increase feature saliency for better point cloud classification. Our SE-PointNet++ architecture has been evaluated on the Titan multispectral LiDAR test datasets and achieved an overall accuracy, a mean Intersection over Union (mIoU), an F1-score, and a Kappa coefficient of 91.16%, 60.15%, 73.14%, and 0.86, respectively. Comparative studies with five established deep learning models confirmed that our proposed SE-PointNet++ achieves promising performance in multispectral LiDAR point cloud classification tasks.

This paper presents a rotation-invariant method for detecting aeroplanes from high-resolution remotely sensed images. First, a superpixel-based strategy is proposed to generate salient and ...

Our work addresses the problem of classifying tree species from mobile LiDAR data. The work is a two step-wise strategy, including tree segmentation and tree species classification. In the tree segmentation step, a voxel-based upward growing filtering is proposed to remove terrain points from the mobile laser scanning data. Then, individual trees are segmented via a Euclidean distance clustering approach and Voxel-based Normalized Cut (VNCut) segmentation approach. In the tree species classification, a voxel-based 3D convolutional neural network (3D-CNN) model is developed based on intensity information. A road section data acquired by a RIEGL VMX-450 system are selected for evaluating the proposed tree classification method. Qualitative analysis shows that our algorithm achieves a good performance.

Multispectral LiDAR, characterization of completeness, and consistency of spectrum and spatial geometric data provide a new data source for land cover classification. However, how to choose the optimal features for a given set of land covers is an open problem for effective land cover classification. To address this problem, we propose a comparative scheme, which investigates a popular deep learning (deep Boltzmann machine, DBM) model for high-level feature representation and widely used machine learning methods for low-level feature extraction and selection [principal component analysis (PCA) and random forest (RF)] in land cover classification. The comparative study was conducted on the multispectral LiDAR point clouds, acquired by a Teledyne Optech's Titan airborne system. The deep learning-based high-level feature representation experimental results showed that, on an ordinary personal computer or workstation, this method required larger training samples and more computational complexity than the machine learning-based low-level feature extraction and selection methods. However, our comparative experiments demonstrated that the classification accuracies of the DBM-based method were higher than those of the RF-based and PCA-based methods using multispectral LiDAR data.

The up-to-date and information-accurate road database plays a significant role in many applications. Recently, with the improvement in image resolutions and quality, remote sensing images have provided an important data source for road extraction tasks. However, due to the topology variations, spectral diversities, and complex scenarios, it is still challenging to realize fully automated and highly accurate road extractions from remote sensing images. This paper proposes a novel dual-attention capsule U-Net (DA-CapsUNet) for road region extraction by combining the advantageous properties of capsule representations and the powerful features of attention mechanisms. By constructing a capsule U-Net architecture, the DA-CapsUNet can extract and fuse multiscale capsule features to recover a high-resolution and semantically strong feature representation. By designing the multiscale context-augmentation and two types of feature attention modules, the DA-CapsUNet can exploit multiscale contextual properties at a high-resolution perspective and generate an informative and class-specific feature encoding. Quantitative evaluations on a large dataset showed that the DA-CapsUNet provides a competitive road extraction performance with a precision of 0.9523, a recall of 0.9486, and an F-score of 0.9504, respectively. Comparative studies with eight recently developed deep learning methods also confirmed the applicability and superiority or compatibility of the DA-CapsUNet in road extraction tasks.

This paper presents a traffic sign detection and recognition method from mobile LiDAR data and digital images for intelligent transportation-related applications. The traffic sign detection and recognition method includes two steps: traffic sign interest regions are first extracted from mobile LiDRA data. Next, traffic signs are identified from digital images simultaneously collected from the multi-sensor mobile LiDAR systems via a convolutional capsule network model. The experimental results demonstrate that the proposed method obtains a promising, reliable, and high performance in both detecting traffic signs in 3-D point clouds and recognizing traffic signs on 2-D images.

As a kind of very large scale integrated circuit, the surface temperature of CPU is closely affected by its utilization. However, there relatively lacks quantitative analysis on the relationship between the surface temperature and CPU utilization at present. In this study the CPU utilization is controlled by a load tool running on the kernels, and the number of kernels get into operation is gradually increased from the initial to the steady-state threshold, meanwhile the surface temperature of CPU die without thermal protection is collected based on the infrared testing technology. Then the utilization rates of each core under different load levels are obtained, and the model coefficients about the surface temperature and CPU utilization are fitted according to the data collected, so that the quantitative relationship between the surface temperature and CPU utilization is analyzed.

Unmanned aerial vehicles using light detection and ranging (UAV LiDAR) with high spatial resolution have shown great potential in forest applications because they can capture vertical structures of forests. Individual tree segmentation is the foundation of many forest research works and applications. The tradition fixed bandwidth mean shift has been applied to individual tree segmentation and proved to be robust in tree segmentation. However, the fixed bandwidth-based segmentation methods are not suitable for various crown sizes, resulting in omission or commission errors. Therefore, to increase tree-segmentation accuracy, we propose a self-adaptive bandwidth estimation method to estimate the optimal kernel bandwidth automatically without any prior knowledge of crown size. First, from the global maximum point, we divide the three-dimensional (3D) space into a set of angular sectors, for each of which a canopy surface is simulated and the potential tree crown boundaries are identified to estimate average crown width as the kernel bandwidth. Afterwards, we use a mean shift with the automatically estimated kernel bandwidth to extract individual tree points. The method is iteratively implemented within a given area until all trees are segmented. The proposed method was tested on the 7 plots acquired by a Velodyne 16E LiDAR system, including 3 simple plots and 4 complex plots, and 95% and 80% of trees were correctly segmented, respectively. Comparative experiments show that our method contributes to the improvement of both segmentation accuracy and computational efficiency.

Using mobile Light Detection and Ranging point clouds to accomplish road scene labeling tasks shows promise for a variety of applications. Most existing methods for semantic labeling of point clouds require a huge number of fully supervised point cloud scenes, where each point needs to be manually annotated with a specific category. Manually annotating each point in point cloud scenes is labor intensive and hinders practical usage of those methods. To alleviate such a huge burden of manual annotation, in this paper, we introduce an active learning method that avoids annotating the whole point cloud scenes by iteratively annotating a small portion of unlabeled supervoxels and creating a minimal manually annotated training set. In order to avoid the biased sampling existing in traditional active learning methods, a neighbor-consistency prior is exploited to select the potentially misclassified samples into the training set to improve the accuracy of the statistical model. Furthermore, lots of methods only consider short-range contextual information to conduct semantic labeling tasks, but ignore the long-range contexts among local variables. In this paper, we use a higher order Markov random field model to take into account more contexts for refining the labeling results, despite of lacking fully supervised scenes. Evaluations on three data sets show that our proposed framework achieves a high accuracy in labeling point clouds although only a small portion of labels is provided. Moreover, comparative experiments demonstrate that our proposed framework is superior to traditional sampling methods and exhibits comparable performance to those fully supervised models.

This study aims at building a robust method for detecting and classifying traffic signs from mobile LiDAR point clouds and digital images. First, this method detects traffic signs from mobile LiDAR point clouds with regard to a prior knowledge of road width, pole height, reflectance, geometrical structure, and traffic-sign size. Then, traffic-sign images are segmented by projecting the detected traffic-sign points onto the digital images. Afterward, the segmented traffic-sign images are normalized for automatic classification with a given image size. Finally, a traffic-sign classifier is proposed based on a supervised Gaussian–Bernoulli deep Boltzmann machine model. We evaluated the proposed method using datasets acquired by a RIEGL VMX-450 system. The traffic-sign detection accuracy of 86.8% was achieved; through parameter sensitivity analysis, the overall performance of traffic-sign classification achieved a recognition rate of 93.3%. The computational performance showed that our method provides a promising solution to traffic-sign detection and classification using mobile LiDAR point clouds and digital images.

This article describes how text documents are a major data structure in the era of big data. With the explosive growth of data, the number of documents with multi-labels has increased dramatically. The popular multi-label classification technology, which is usually employed to handle multinomial text documents, is sensitive to the noise terms of text documents. Therefore, there still exists a huge room for multi-label classification of text documents. This article introduces a supervised topic model, named labeled LDA with function terms (LF-LDA), to filter out the noisy function terms from text documents, which can help to improve the performance of multi-label classification of text documents. The article also shows the derivation of the Gibbs Sampling formulas in detail, which can be generalized to other similar topic models. Based on the textual data set RCV1-v2, the article compared the proposed model with other two state-of-the-art multi-label classifiers, Tuned SVM and labeled LDA, on both Macro-F1 and Micro-F1 metrics. The result shows that LF-LDA outperforms them and has the lowest variance, which indicates the robustness of the LF-LDA classifier.

As the optical remote sensing techniques keep developing with a rapid pace, remote sensing images are positively considered in many fields. Accordingly, a great number of algorithms have been exploited for remote sensing image interpretation purposes. Thereinto, object recognition acts as an important ingredient to many applications. However, to achieve highly accurate object recognition is still challengeable caused by the orientation and size diversities, spatial distribution and density variations, shape and aspect ratio irregularities, occlusion and shadow impacts, as well as complex texture and surrounding environment changes. In this paper, a sparse anchoring guided high-resolution capsule network (SAHR-CapsNet) is proposed for geospatial object detection based on remote sensing images. First, formulated with the multibranch high-resolution capsule network architecture assisted by multiscale feature propagation and fusion, the SAHR-CapsNet can extract semantically strong and spatially accurate feature semantics at multiple scales. Second, integrated with the efficient capsule-based self-attention module, the SAHR-CapsNet functions promisingly to attend to target-specific spatial features and informative channel features. Finally, adopted with the capsule-based sparse anchoring network, the SAHR-CapsNet performs efficiently in generating a fixed number of lightweight, high-quality sparse region proposals. Quantitative assessments and comparative analyses on two challenging remote sensing image datasets demonstrate the applicability and effectiveness of the developed SAHR-CapsNet for geospatial object detection applications.

With the explosive growth of academic writing, it is difficult for researchers to find significant papers in their area of interest. In this paper, we propose a pipeline model, named collective topical PageRank, to evaluate the topic-dependent impact of scientific papers. First, we fit the model to a correlation topic model based on the textual content of papers to extract scientific topics and correlations. Then, we present a modified PageRank algorithm, which incorporates the venue, the correlations of the scientific topics, and the publication year of each paper into a random walk to evaluate the paper’s topic-dependent academic impact. Our experiments showed that the model can effectively identify significant papers as well as venues for each scientific topic, recommend papers for further reading or citing, explore the evolution of scientific topics, and calculate the venues’ dynamic topic-dependent academic impact.

At present, the incidence rate of colorectal cancer (CRC) is increasing year by year. It has always affected people's physical and mental health and quality of life. How to improve the detection ability of polyp plays a key role in colonoscopy. In order to solve these problems, in this paper, we first enhance the contrast of the input image by well distinguishing the foreground from the background in order to improve the saliency of the polyp regions. Then, we feed the enhanced image into an improved Faster R-CNN architecture comprised three processing modules for feature extraction, region proposal generation, and polyp detection, respectively. In order to further improve the quality, as well as the feature abstraction capability of the feature maps produced by the feature extraction network, we append an attention module to attend to the useful feature channels and weaken the contributions of the helpless feature channels. The experimental results demonstrate that the accuracy of the proposed polyp detection network is greatly improved compared with the existing algorithms, and the network not only can accurately identify polyps of varying sizes and conditions in single polyp images, but also can achieve excellent performance in handling multiple polyp images. This paper will be greatly helpful to alleviate the missed diagnosis of clinicians in the process of endoscopic examination and disease treatment, as well as providing effective assistance for the early diagnosis, treatment and prevention of the CRC, which is also of great significance to the clinical work of physicians.

Rolling bearing’s fault mode usually shows compound faults in aero-engine. The compound faults characteristics are more complex than single one, and many signal analysis methods have rather great limitation for compound fault characteristic extraction which leads to the difficulty to monitor the running state of rolling bearing in aero-engine. Based on above analysis, a method of combining wavelet transform with cyclostationary theory, autocorrelation function and Hilbert transform is proposed and applied to extract characteristic frequency of rolling bearing from compound faults mode only according to single-channel vibration acceleration signal of aero-engine. Meanwhile, a consideration is given to the influence of sensor installation position, compound fault types in the extraction of compound faults characteristics. The result indicates that the proposed new method can effectively monitor rolling bearing running state in four different compound fault modes just according to single-channel vibration acceleration signal no matter sensors are installed in horizontal or vertical direction.

This letter proposes a processing chain for detecting aeroplanes from very high-resolution (VHR) remotely sensed images with the fusion of deep feature representation and rotation-invariant Hough forests. First, superpixel segmentation is used to generate meaningful and non-redundant patches. Second, deep learning techniques are exploited to construct a multi-layer feature encoder for representing high-order features of patches. Third, a set of multi-scale rotation-invariant Hough forests are trained to detect aeroplanes of varying orientations and sizes. Experiments show that the proposed method is a promising solution for detecting aeroplanes from VHR remotely sensed images, with a completeness, correctness, and F-measure of 0.956, 0.970, and 0.963, respectively. Comparative studies with four existing methods also demonstrate that the proposed method outperforms the other existing methods in accurately detecting aeroplanes of varying appearances, orientations, and sizes.

This paper presents a supervoxel-based approach for automated localization and extraction of street light poles in point clouds acquired by a mobile LiDAR system. The method consists of five steps: preprocessing, localization, segmentation, feature extraction, and classification. First, the raw point clouds are divided into segments along the trajectory, the ground points are removed, and the remaining points are segmented into supervoxels. Then, a robust localization method is proposed to accurately identify the pole-like objects. Next, a localization-guided segmentation method is proposed to obtain pole-like objects. Subsequently, the pole features are classified using the support vector machine and random forests. The proposed approach was evaluated on three datasets with 1,055 street light poles and 701 million points. Experimental results show that our localization method achieved an average recall value of 98.8%. A comparative study proved that our method is more robust and efficient than other existing methods for localization and extraction of street light poles.

Advances in information technologies have brought many changes to our lives. Finding free music through online platform rather than buying hard copies offline is one of the most significant changes. Recent researches suggest that as users becoming more engaged with the online content-provider platform, they are more willing to pay for the service or premium service. This study addresses the need of freemium e-commerce identifying potential subscribers. In specific, we propose a novel method, namely support vector classification (SVC), to categorize content viewers into potential subscriber and non-potential subscribers. Our method provides satisfied prediction result and the experiment showed that SVC is a superior method in this kind of task.

This paper proposes a novel algorithm for detecting road scene objects (e.g., light poles, traffic signposts, and cars) from 3-D mobile-laser-scanning point cloud data for transportation-related applications. To describe local abstract features of point cloud objects, a contextual visual vocabulary is generated by integrating spatial contextual information of feature regions. Objects of interest are detected based on the similarity measures of the bag of contextual-visual words between the query object and the segmented semantic objects. Quantitative evaluations on two selected data sets show that the proposed algorithm achieves an average recall, precision, quality, and F-score of 0.949, 0.970, 0.922, and 0.959, respectively, in detecting light poles, traffic signposts, and cars. Comparative studies demonstrate the superior performance of the proposed algorithm over other existing methods.

Vehicle detection from remote sensing images plays a significant role in transportation related applications. However, the scale variations, orientation variations, illumination variations, and partial occlusions of vehicles, as well as the image qualities, bring great challenges for accurate vehicle detection. In this paper, we present an affine-function transformation-based object matching framework for vehicle detection from unmanned aerial vehicle (UAV) images. First, meaningful and non-redundant patches are generated through a superpixel segmentation strategy. Then, the affine-function transformation-based object matching framework is applied to a vehicle template and each of the patches for vehicle existence estimation. Finally, vehicles are detected and located after matching cost thresholding, vehicle location estimation, and multiple response elimination. Quantitative evaluations on two UAV image datasets show that the proposed method achieves an average completeness, correctness, quality, and F1-measure of 0.909, 0.969, 0.883, and 0.938, respectively. Comparative studies also demonstrate that the proposed method achieves compatible performance with the Faster R-CNN and outperforms the other eight existing methods in accurately detecting vehicles of various conditions.

Forests play a key role in terrestrial ecosystems, and the variables extracted from single trees can be used in various fields and applications for evaluating forest production and assessing forest ecosystem services. In this study, we developed an automated hierarchical single-tree segmentation approach based on the high density three-dimensional (3D) Unmanned Aerial Vehicle (UAV) point clouds. First, this approach obtains normalized non-ground UAV points in data preprocessing; then, a voxel-based mean shift algorithm is used to roughly classify the non-ground UAV points into well-detected and under-segmentation clusters. Moreover, potential tree apices for each under-segmentation cluster are obtained with regard to profile shape curves and finally input to the normalized cut segmentation (NCut) algorithm to segment iteratively the under-segmentation cluster into single trees. We evaluated the proposed method using datasets acquired by a Velodyne 16E LiDAR system mounted on a multi-rotor UAV. The results showed that the proposed method achieves the average correctness, completeness, and overall accuracy of 0.90, 0.88, and 0.89, respectively, in delineating single trees. Comparative analysis demonstrated that our method provided a promising solution to reliable and robust segmentation of single trees from UAV LiDAR data with high point cloud density.

This paper presents an automated algorithm for rapidly and effectively detecting cars directly from large-volume 3-D point clouds. Rather than using low-order descriptors, a multilayer feature generation model is created to obtain high-order feature representations for 3-D local patches through deep learning techniques. To handle cars with different levels of incompleteness caused by data acquisition ways and occlusions, a hierarchical visibility estimation model is developed to augment Hough voting. Considering scale and orientation variations in the azimuth direction, a set of multiscale Hough forests is constructed to rotationally cast votes to estimate cars' centroids. Quantitative assessments show that the proposed algorithm achieves average completeness, correctness, quality, and $F_{1}$ -measure of 0.94, 0.96, 0.90, and 0.95, respectively, in detecting 3-D cars. Comparative studies also demonstrate that the proposed algorithm outperforms the other four existing algorithms in accurately and completely detecting 3-D cars from large-scale 3-D point clouds.

Semantic labeling of road scenes using colorized mobile LiDAR point clouds is of great significance in a variety of applications, particularly intelligent transportation systems. However, many challenges, such as incompleteness of objects caused by occlusion, overlapping between neighboring objects, interclass local similarities, and computational burden brought by a huge number of points, make it an ongoing open research area. In this paper, we propose a novel patch-based framework for labeling road scenes of colorized mobile LiDAR point clouds. In the proposed framework, first, three-dimensional (3-D) patches extracted from point clouds are used to construct a 3-D patch-based match graph structure (3D-PMG), which transfers category labels from labeled to unlabeled point cloud road scenes efficiently. Then, to rectify the transferring errors caused by local patch similarities in different categories, contextual information among 3-D patches is exploited by combining 3D-PMG with Markov random fields. In the experiments, the proposed framework is validated on colorized mobile LiDAR point clouds acquired by the RIEGL VMX-450 mobile LiDAR system. Comparative experiments show the superior performance of the proposed framework for accurate semantic labeling of road scenes.

This letter addresses the problem of detecting pole-like road objects (including light poles and traffic signposts) from mobile light detection and ranging (LiDAR) data for transportation-related applications. The method consists of two consecutive stages: training and pole-like object detection. At the training stage, a contextual visual vocabulary is created from the feature regions generated from a training data set by supervoxel segmentation. At the pole-like object detection stage, a bag-of-contextual-visual-words representation is generated for each semantic object segmented from mobile LiDAR data. The experimental results show that the proposed method achieves correctness, omission, and commission of 88.9%, 11.1%, and 2.8%, respectively, in detecting pole-like road objects. Computational complexity analysis demonstrates that our method provides a promising and effective solution to rapid and accurate detection of pole-like objects from large volumes of mobile LiDAR data.

This paper presents a novel algorithm for detection and recognition of traffic signs in mobile laser scanning (MLS) data for intelligent transportation-related applications. The traffic sign detection task is accomplished based on 3-D point clouds by using bag-of-visual-phrases representations; whereas the recognition task is achieved based on 2-D images by using a Gaussian-Bernoulli deep Boltzmann machine-based hierarchical classifier. To exploit high-order feature encodings of feature regions, a deep Boltzmann machine-based feature encoder is constructed. For detecting traffic signs in 3-D point clouds, the proposed algorithm achieves an average recall, precision, quality, and F-score of 0.956, 0.946, 0.907, and 0.951, respectively, on the four selected MLS datasets. For on-image traffic sign recognition, a recognition accuracy of 97.54% is achieved by using the proposed hierarchical classifier. Comparative studies with the existing traffic sign detection and recognition methods demonstrate that our algorithm obtains promising, reliable, and high performance in both detecting traffic signs in 3-D point clouds and recognizing traffic signs on 2-D images.

This paper proposes a rotation-and-scale-invariant method for detecting airplanes from high-resolution satellite images. To improve feature representation capability, a multi-layer feature generation model is created to produce high-order feature representations for local image patches through deep learning techniques. To effectively estimate airplane centroids, a Hough forest model is trained to learn mappings from high-order patch features to the probabilities of an airplane being present at specific locations. To handle airplanes with varying orientations, patch orientation is defined and integrated into the Hough forest to augment Hough voting. The scale invariance is achieved by using a set of scale factors embedded in the Hough forest. Quantitative evaluations on the images collected from Google Earth service show that the proposed method achieves a completeness, correctness, quality, and F1-measure of 0.968, 0.972, 0.942, and 0.970, respectively, in detecting airplanes with arbitrary orientations and sizes. Comparative studies also demonstrate that the proposed method outperforms the other three existing methods in accurately and completely detecting airplanes in high-resolution remotely sensed images.

This paper proposes a novel framework for automated detection of urban road manhole covers using mobile laser scanning (MLS) data. First, to narrow searching regions and reduce the computational complexity, road surface points are segmented from a raw point cloud via a curb-based road surface segmentation approach and rasterized into a georeferenced intensity image through inverse distance weighted interpolation. Then, a supervised deep learning model is developed to construct a multilayer feature generation model for depicting high-order features of local image patches. Next, a random forest model is trained to learn mappings from high-order patch features to the probabilities of the existence of urban road manhole covers centered at specific locations. Finally, urban road manhole covers are detected from georeferenced intensity images based on the multilayer feature generation model and random forest model. Quantitative evaluations show that the proposed algorithm achieves an average completeness, correctness, quality, and $\mathrm{F}_1$ -measure of 0.955, 0.959, 0.917, and 0.957, respectively, in detecting urban road manhole covers from georeferenced intensity images. Comparative studies demonstrate the advantageous performance of the proposed algorithm over other existing methods for rapid and automated detection of urban road manhole covers using MLS data.