HOG-VGG:VGG Network with HOG Feature Fusion for High-Precision PolSAR Terrain Classification

This article proposes a VGG network with histogram of oriented gradient(HOG) feature fusion(HOG-VGG) for polarization synthetic aperture radar(PolSAR) image terrain classification.VGG-Net has a strong ability of deep feature extraction,which can fully extract the global deep features of different terrains in PolSAR images,so it is widely used in PolSAR terrain classification.However,VGG-Net ignores the local edge & shape features,resulting in incomplete feature representation of the PolSAR terrains,as a consequence,the terrain classification accuracy is not promising.In fact,edge and shape features play an important role in PolSAR terrain classification.To solve this problem,a new VGG network with HOG feature fusion was specifically proposed for high-precision PolSAR terrain classification.HOG-VGG extracts both the global deep semantic features and the local edge & shape features of the PolSAR terrains,so the terrain feature representation completeness is greatly elevated.Moreover,HOG-VGG optimally fuses the global deep features and the local edge & shape features to achieve the best classification results.The superiority of HOG-VGG is verified on the Flevoland,San Francisco and Oberpfaffenhofen datasets.Experiments show that the proposed HOG-VGG achieves much better PolSAR terrain classification performance,with overall accuracies of 97.54%,94.63%,and 96.07%,respectively.

Weighted Multi-sensor Data Level Fusion Method of Vibration Signal Based on Correlation Function

As the differences of sensor＇s precision and some random factors are difficult to control,the actual measurement signals are far from the target signals that affect the reliability and precision of rotating machinery fault diagnosis.The traditional signal processing methods,such as classical inference and weighted averaging algorithm usually lack dynamic adaptability that is easy for trends to cause the faults to be misjudged or left out.To enhance the measuring veracity and precision of vibration signal in rotary machine multi-sensor vibration signal fault diagnosis,a novel data level fusion approach is presented on the basis of correlation function analysis to fast determine the weighted value of multi-sensor vibration signals.The approach doesn＇t require knowing the prior information about sensors,and the weighted value of sensors can be confirmed depending on the correlation measure of real-time data tested in the data level fusion process.It gives greater weighted value to the greater correlation measure of sensor signals,and vice versa.The approach can effectively suppress large errors and even can still fuse data in the case of sensor failures because it takes full advantage of sensor＇s own-information to determine the weighted value.Moreover,it has good performance of anti-jamming due to the correlation measures between noise and effective signals are usually small.Through the simulation of typical signal collected from multi-sensors,the comparative analysis of dynamic adaptability and fault tolerance between the proposed approach and traditional weighted averaging approach is taken.Finally,the rotor dynamics and integrated fault simulator is taken as an example to verify the feasibility and advantages of the proposed approach,it is shown that the multi-sensor data level fusion based on correlation function weighted approach is better than the traditional weighted average approach with respect to fusion precision and dynamic adaptability.Meantime,the approach is adaptable and easy to use,can be applied to other areas of vibration measurement.

Four Levels Anterior Cervical Discectomy and Fusion by Stand Alone PEEK Cages

Background: cervical spondylotic myelopathy is a common health problem that neurosurgeons face in Egypt. The aim of this study is to evaluate the efficacy of PEEK cage only in 4 levels anterior cervical discectomy as one of surgical option other than anterior cervical corpectomy, fixation by plat or posterior approach for cervical laminectomy, and assessment of post spinal surgery pain. Methods: this prospective study on 28 patients with cervical spondylotic myelopathy (CSM) over a period of 3 years (between April 2012 and April 2015) with mean period of follow up 30 months. We have done anterior cervical discectomy with fixation by cage only for all cases with perioperative assessment and scoring clinically and radiologically (Japanese Orthopaedic Association [JOA] scores, Visual Analogue Scale [VAS] scores for assessment of neck and arm pain, perioperative parameters (hospital stay, blood loss, operative time), the European Myelopathy Scoring (EMS) and Odom’s criteria, and the incidence of complication,post spinal surgery pain assessment). Results: clinical outcome was excellent (28.55), good (50%) and fair (21.5) according to Odom criteria. The European Myelopathy Scoring (EMS), improved from 10 to 16. The mean JOA score improved from 10.1 ± 2.1 to 14.2 ± 2.3. Fusion failure had been seen in 4 patients in one level for each secondary to anterior displacement of the cage with no other major complications. Conclusion: 4 levels anterior cervical discectomy with PEEK cage only is an effective, save and less costly with less post operative complication and hospital stay and less post spinal surgery pain.

A Local Contrast Fusion Based 3D Otsu Algorithm for Multilevel Image Segmentation

To overcome the shortcomings of 1 D and 2 D Otsu’s thresholding techniques, the 3 D Otsu method has been developed.Among all Otsu’s methods, 3 D Otsu technique provides the best threshold values for the multi-level thresholding processes. In this paper, to improve the quality of segmented images, a simple and effective multilevel thresholding method is introduced. The proposed approach focuses on preserving edge detail by computing the 3 D Otsu along the fusion phenomena. The advantages of the presented scheme include higher quality outcomes, better preservation of tiny details and boundaries and reduced execution time with rising threshold levels. The fusion approach depends upon the differences between pixel intensity values within a small local space of an image;it aims to improve localized information after the thresholding process. The fusion of images based on local contrast can improve image segmentation performance by minimizing the loss of local contrast, loss of details and gray-level distributions. Results show that the proposed method yields more promising segmentation results when compared to conventional1 D Otsu, 2 D Otsu and 3 D Otsu methods, as evident from the objective and subjective evaluations.

Machine Learning for Data Fusion:A Fuzzy AHP Approach for Open Issues

Data fusion generates fused data by combining multiple sources,resulting in information that is more consistent,accurate,and useful than any individual source and more reliable and consistent than the raw original data,which are often imperfect,inconsistent,complex,and uncertain.Traditional data fusion methods like probabilistic fusion,set-based fusion,and evidential belief reasoning fusion methods are computationally complex and require accurate classification and proper handling of raw data.Data fusion is the process of integrating multiple data sources.Data filtering means examining a dataset to exclude,rearrange,or apportion data according to the criteria.Different sensors generate a large amount of data,requiring the development of machine learning(ML)algorithms to overcome the challenges of traditional methods.The advancement in hardware acceleration and the abundance of data from various sensors have led to the development of machine learning(ML)algorithms,expected to address the limitations of traditional methods.However,many open issues still exist as machine learning algorithms are used for data fusion.From the literature,nine issues have been identified irrespective of any application.The decision-makers should pay attention to these issues as data fusion becomes more applicable and successful.A fuzzy analytical hierarchical process(FAHP)enables us to handle these issues.It helps to get the weights for each corresponding issue and rank issues based on these calculated weights.The most significant issue identified is the lack of deep learning models used for data fusion that improve accuracy and learning quality weighted 0.141.The least significant one is the cross-domain multimodal data fusion weighted 0.076 because the whole semantic knowledge for multimodal data cannot be captured.

Vertical deformation analysis based on combined adjustment for GNSS and leveling data

A method is proposed to fuse the velocity data of the global navigation satellite system(GNSS) and leveling height via combined adjustment with constraints. First, stable GNSS-leveling points are uniformly selected, and the constraints of the geodetic height change velocity and normal height change velocity are given. Then, the GNSS vertical velocities and leveling height difference are used as observations of combined adjustment, and robust least-squares estimation are used to estimate the velocities of the unknown points. Finally, a vertical movement model is established with the GNSS vertical velocities and leveling vertical velocities obtained via combined adjustment. Data from the second-order leveling network and GNSS control points in Shandong Province are taken as test data, and eight calculation schemes are used for discussion. One of the schemes, the bifactor robust combined adjustment method based on variance component estimation with two kinds of vertical velocity constraints achieves the optimal results. The method applied in the scheme can be recommended for data fusion of GNSS and leveling, further improving the reliability of vertical crustal movement in Shandong Province.

Enhanced Feature Fusion Segmentation for Tumor Detection Using Intelligent Techniques

In thefield of diagnosis of medical images the challenge lies in tracking and identifying the defective cells and the extent of the defective region within the complex structure of a brain cavity.Locating the defective cells precisely during the diagnosis phase helps tofight the greatest exterminator of mankind.Early detec-tion of these defective cells requires an accurate computer-aided diagnostic system(CAD)that supports early treatment and promotes survival rates of patients.An ear-lier version of CAD systems relies greatly on the expertise of radiologist and it con-sumed more time to identify the defective region.The manuscript takes the efficacy of coalescing features like intensity,shape,and texture of the magnetic resonance image(MRI).In the Enhanced Feature Fusion Segmentation based classification method(EEFS)the image is enhanced and segmented to extract the prominent fea-tures.To bring out the desired effect the EEFS method uses Enhanced Local Binary Pattern(EnLBP),Partisan Gray Level Co-occurrence Matrix Histogram of Oriented Gradients(PGLCMHOG),and iGrab cut method to segment image.These prominent features along with deep features are coalesced to provide a single-dimensional fea-ture vector that is effectively used for prediction.The coalesced vector is used with the existing classifiers to compare the results of these classifiers with that of the gen-erated vector.The generated vector provides promising results with commendably less computatio nal time for pre-processing and classification of MR medical images.

多层次时空特征自适应集成与特有-共享特征融合的双模态情感识别

在结合脑电(EEG)信号与人脸图像的双模态情感识别领域中,通常存在两个挑战性问题:(1)如何从EEG信号中以端到端方式学习到更具显著性的情感语义特征;(2)如何充分利用双模态信息,捕捉双模态特征中情感语义的一致性与互补性。为此,提出了多层次时空特征自适应集成与特有-共享特征融合的双模态情感识别模型。一方面,为从EEG信号中获得更具显著性的情感语义特征,设计了多层次时空特征自适应集成模块。该模块首先通过双流结构捕捉EEG信号的时空特征,再通过特征相似度加权并集成各层次的特征,最后利用门控机制自适应地学习各层次相对重要的情感特征。另一方面,为挖掘EEG信号与人脸图像之间的情感语义一致性与互补性,设计了特有-共享特征融合模块,通过特有特征的学习和共享特征的学习来联合学习情感语义特征,并结合损失函数实现各模态特有语义信息和模态间共享语义信息的自动提取。在DEAP和MAHNOB-HCI两种数据集上,采用跨实验验证和5折交叉验证两种实验手段验证了提出模型的性能。实验结果表明,该模型取得了具有竞争力的结果,为基于EEG信号与人脸图像的双模态情感识别提供了一种有效的解决方案。

多模态方面级情感分析的多视图交互学习网络

以往的多模态方面级情感分析方法只利用预训练模型的一般文本和图片表示,对方面和观点词相关性的识别不敏感,且不能动态获取图片信息对单词表示的贡献,因而不能充分识别多模态与方面之间的相关性。针对上述问题,提出一种多视图交互学习网络模型。将句子从上下文和句法两个视图上分别提取特征,以便在多模态交互时充分利用到文本的全局特征;对文本、图片和方面之间的关系进行建模,使模型实现多模态交互;同时融合不同模态的交互表示,动态获取视觉信息对文本中每个单词的贡献程度,充分提取模态与方面之间的相关性。最后通过全连接层和Softmax层获取情感分类结果。在两个数据集上进行实验,实验结果表明该模型能够有效增强多模态方面级情感分类的效果。

基于排水系统主动降水位的污涝同治方案设计

合流制排水系统具有投资少、工期短、见效快等优点,但也容易导致污水管网高水位运行,反噬下游地区排水系统安全,导致旱季污水溢流入河污染水环境,雨季内涝积水影响水安全。通过研究污水泵站进水管水位变化,查明了D片区污涝矛盾突出的原因。采取熔断措施构建独立排水区域后,排水管网水位降低1.6 m以上,开闸次数由83次/年降低到17次/年,COD入河量减少248.4 t/年,氨氮19.7 t/年,保障河道水质达到Ⅴ类。

基于改进Grabcut分割与多特征决策融合的电力线放电痕迹识别

电力线触树故障中,导线表面的遗留痕迹是事故防治和责任认定的重要依据,但目前中外针对触树后电力线放电痕迹特征规律和辨识方法的研究极其匮乏。为此,搭建10 kV中压线路触树放电实验平台,采集放电后的导线表面痕迹图像,并对导线表面痕迹特征进行系统分析,为人工巡检和智能化痕迹识别提供基础依据。然后,提出改进型Grabcut前景提取方法,综合利用U^(2)Net的自动分割特点和Grabcut的高精度优势,解决Grabcut算法中初始框无法自动确定的问题,实现复杂背景下导线痕迹区域自动精准分割。最后,提出基于低层纹理、颜色特征和高层深度特征的导线表面痕迹全面表征,并采用多数投票规则实现低层和高层特征识别结果决策融合,获得导线痕迹辨识结果,测试实验中平均识别准确率达到91.68%,证明了方法的有效性。

激光雷达和相机的决策级融合目标检测方法

【目的】激光雷达与相机这两类传感器检测数据格式不统一、分辨率不同,且数据级和特征级的融合计算复杂度高,故提出一种决策级的目标融合检测方法。【方法】对激光雷达与相机的安装位置进行联合标定,实现这两类传感器检测结果的坐标系转换;利用匈牙利算法将激光雷达点云检测目标框和相机图像检测目标框进行匹配,设定目标框重合面积阈值,检测获得目标物的位置、类型等。【结果】实车测试结果表明,根据检测目标检测框长宽比选取不同交并比阈值的方法使得车辆和行人的目标识别准确率分别提升了3.3%和5.3%。利用公开数据集KITTI对所提融合方法进行验证,结果表明,在3种不同难度等级场景下,所提融合方法的检测精度分别达到了75.42%、69.71%、63.71%,与现有常用的融合方法相比,检测精度均有所提升。【结论】这两类传感器的检测目标框重合面积阈值对决策级融合检测结果影响较大,根据检测目标检测框长宽比选取不同阈值可有效提升车辆和行人的目标识别准确率。决策级融合方法能准确匹配雷达和相机的检测目标,有效提升目标检测精度。

腰椎融合术后患者急性疼痛发展轨迹及影响因素

目的探讨开放性腰椎融合术患者急性术后疼痛轨迹的潜在类别,分析不同潜在类别的影响因素,为不同类别患者制定针对性镇痛措施提供依据。方法使用便利抽样法,选取择期开放性腰椎融合术患者251例,在术前1 d收集患者一般资料、疼痛评分及睡眠质量资料。术后分别于全麻清醒并拔除气管导管的0 h、6 h、12 h、1 d、2 d、3 d、4 d、5 d、6 d 9个时间点评估患者疼痛程度。采用潜类别增长模型识别急性术后疼痛轨迹的潜在类别,采用无序多分类logistic回归分析急性术后疼痛轨迹潜在类别的影响因素。结果识别出3种急性术后疼痛轨迹,分别为疼痛加剧后缓解组(83例,33.0%),疼痛加剧稳定后反弹组(30例,12.0%),轻度疼痛组(138例,55.0%)。回归分析显示,相对于轻度疼痛组,病程≥10年、高中以下文化程度、手术时间>240 min的患者更容易进入疼痛加剧稳定后反弹组;年龄≤50岁、存在睡眠障碍、术后未使用镇痛泵的患者更容易进入疼痛加剧后缓解组(均P<0.05)。结论开放性腰椎融合手术患者急性术后疼痛呈现不同的发展轨迹,医护人员应根据不同发展轨迹重视年龄≤50岁、病程≥10年、高中以下文化程度、术前有睡眠障碍、手术时间>240 min、术后未使用镇痛泵的患者,给予及时的疼痛评估及针对性干预措施,改善患者术后疼痛程度。

改进YOLOv8的农作物叶片病虫害识别算法

针对传统检测网络难以准确、高效地提取农作物叶片病虫害特征信息的问题,通过改进YOLOv8网络,提出一种多层级多尺度特征融合的农作物叶片病虫害识别算法。通过学习不同层级特征直接的特征关系,构建多层级特征编码模块,学习全面的特征表达;在Transformer的基础上设计多尺度空间—通道注意力模块,利用学习细粒度、粗粒度等多尺度全面的特征表达模式,捕获不同尺度特征之间的互补关系,并将所有特征表示有效融合起来,构成完整的图像特征表示,进而获取更佳的识别结果。在Plant Village公开数据集进行试验验证,结果表明:提出的改进方法能够有效提升配准精度,准确地识别出农作物叶片上同时存在的不同病虫害,对番茄叶片检测的mAP 0.5达到88.74%,比传统YOLOv8方法提升8.53%,且计算耗时没有明显增加。消融试验也充分证明所提各个模块的有效性,能够更好地实现高精度识别叶片病虫害,为农田智慧化管理提供有力支持和保障。

多层次特征融合与超图卷积的生成对抗壁画修复

针对现有壁画深度学习修复方法,存在上下文信息关注不足及结果欠佳的问题,提出了一种多层次特征融合与超图卷积的生成对抗修复模型。首先,利用金字塔特征分层对壁画进行多尺度层次特征提取,并采用混合空洞卷积单元扩大多层特征提取感受野,以克服单尺度卷积操作对于壁画特征提取能力不足的问题。然后,提出多分支短链融合层及门控机制融合多分支特征方法,将相邻分支间的特征信息进行融合,使融合后的壁画特征图中既有同分支的特征,又有相邻分支的特征,以提高特征信息的利用率;并引入门控机制对特征进行选择融合,以减少细节信息的丢失。接着,将融合特征通过卷积长短期记忆网络(ConvLSTM)特征注意力方法,增强对壁画上下文信息的关注。最后,设计超图卷积壁画长程特征增强模块,通过在编码器和解码器的跳跃连接之间建立超图卷积层,利用超图卷积捕获编码器的空间特征信息,并将其迁移到解码器中,有助于解码器更好地生成壁画图像,以加强特征的长程依赖关系,并与SN-PatchGAN判别器对抗博弈从而完成修复。通过敦煌壁画数字化修复实验,结果表明:所提方法客观评价优于对比算法,对于破损壁画修复结果更加清晰自然。

基于特征增强及多层次融合的火灾火焰检测

为提升火灾火焰识别检测方法性能,将传统图像处理与神经网络结合,提出1种基于特征增强及多层次融合的轻量级火灾火焰检测模型。模型利用多种色彩空间转换算法增强火焰特征信息,并设计双阶段多层次特征提取融合结构,配合空间注意力机制对火焰信息由粗到精进行提取;同时,针对火灾火焰特点,引入由浅到深逐步融合的自适应多尺度融合结构,提升对不同阶段火灾目标的检测精度。研究结果表明:本文模型可有效提升火灾火焰的检测效果,且具有更高的稳定性和鲁棒性,可准确高效地实现火灾火焰检测。研究结果可为现有火灾检测设备提供更准确的识别结果,从而更好地预防火灾事故发生。

结合多粒度视图动态融合的多模态方面级情感分析

为了解决以往多模态方面级情感分析研究中存在的特征提取不充分、数据噪声未被有效处理以及多模态数据中的复杂交互被忽视等问题,提出了一种多粒度视图动态融合模型(multi-granularity view dynamic fusion model,MVDFM)。从粗粒度和细粒度两个视角,对文本和图像数据进行向量化编码,以便充分捕捉数据特征,增强模型信息表达能力;提取文本、图像的多粒度视图特征,并设计动态门控自注意力机制,对细粒度级的文本、图像视图进行降噪,进一步保证特征提取质量;为了挖掘不同粒度上多视图之间的互补性和一致性,提出一种三视图分解高阶池化机制,对多粒度视图特征进行两阶段动态融合,得到最终的目标方面词情感极性。实验结果表明,该模型在公共数据集Twitter-2015和Twitter-2017上的准确率和F1值分别达到了78.69%、74.48%以及72.77%、71.61%,相较于最优基线模型分别提升了0.55、0.88个百分点,以及1.67、2.45个百分点。说明该方法能够充分利用多模态数据中包含的深层语义信息,并有效挖掘与目标方面词相关的重要信息,从而提高方面级情感预测效果。

多传感器融合的割草机器人障碍物检测方法

为了提高割草机器人自主作业时的环境感知能力,提出一种基于相机和低成本固态激光雷达融合的障碍物检测方法.首先,基于改进的DBSCAN聚类算法,得到一种可自适应确定聚类参数的KANN-DBSCAN算法,利用该算法对固态激光雷达采集到的三维点云聚类分析,得到障碍物点云并通过相机和固态激光雷达联合标定结果投影到二维图像上;其次,基于SSD目标检测网络完成障碍物样本训练,并对图像信息进行检测识别;最后,为了避免因光线不足或远距离雷达点云稀疏聚类困难,导致视觉或雷达检测性能受限,提出一种优势互补的目标级信息融合策略.试验结果表明:所提信息融合策略在融合两传感器检测结果的基础上,有效避免了环境条件改变单一传感器检测性能受限时,环境感知出现的漏检和误检,经过信息融合后的障碍物综合检出率约为87.5%,相较于单一传感器具有很大的提高,使环境感知信息更加全面可靠.

融合项目特征级信息的稀疏兴趣网络序列推荐

在以往提取多兴趣嵌入的序列推荐模型中仅能通过聚类的方法发现少量兴趣概念,忽视项目交互序列中特征级信息对最终推荐结果的影响。针对此问题,对传统的多兴趣序列推荐模型进行改进,提出一种融合项目特征级信息的稀疏兴趣网络序列推荐模型。实验结果表明,相比其它模型,该模型可以更好捕捉用户的多样化偏好并缓解冷启动问题。在给定数据集上,该模型比传统的序列推荐模型在命中率上平均提高了6.4%,归一化折损累计增益平均提高了8.7%。