In Vitro Fertilization in Maize and Tobacco: The Fusion Patterns During Sexual Cell Fusion

Large scale of sexual cell fusion was carried out in order to observe the fusion pattern during in vitro fertilization with special interest in the influence of cell volume on the membrane behavior. Three patterns could be recognized in sexual cell fusion, which was supported by fluorescent microscopy with the aid of video enhanced system and cooled CCD. It was found that the fusion pattern and cell membrane behavior were related to the cell volume ratio of two fusing cells. The results reported here might be useful for interpreting and evaluating data from in vitro fertilization experiments and for explaining male cytoplasm exclusion during fertilization.

Prediction of Enthalpies of Fusion for Divalent Rare Earth Halides Based on Modeling by Artificial Neural Networks and Pattern Recognition

The artificial neural network (ANN) and the pattern recognition were applied to study the correlation of enthalpies of fusion for divalent rare earth halides with their microstructural parameters,such as ionic radius and electronegativity. The model,represented by a back-propagation netal network, was trained with a 12 set of published data for divalent rare earth halides and then was used to predict the unknown ones. Also the criterion equations were ptesented to determine the enthalpies of fuSion for divalent rare earth halides using pattern recognition in mis work. The results from the model in ANN and criterion equations are in very good agreement with reference data.

Hierarchical particle filter tracking algorithm based on multi-feature fusion

A hierarchical particle filter（HPF） framework based on multi-feature fusion is proposed.The proposed HPF effectively uses different feature information to avoid the tracking failure based on the single feature in a complicated environment.In this approach,the Harris algorithm is introduced to detect the corner points of the object,and the corner matching algorithm based on singular value decomposition is used to compute the firstorder weights and make particles centralize in the high likelihood area.Then the local binary pattern（LBP） operator is used to build the observation model of the target based on the color and texture features,by which the second-order weights of particles and the accurate location of the target can be obtained.Moreover,a backstepping controller is proposed to complete the whole tracking system.Simulations and experiments are carried out,and the results show that the HPF algorithm with the backstepping controller achieves stable and accurate tracking with good robustness in complex environments.

Intent Pattern Recognition of Lower-limb Motion Based on Mechanical Sensors

Based on the regularity nature of lower-limb motion,an intent pattern recognition approach for above-knee prosthesis is proposed in this paper. To remedy the defects of recognizer based on electromyogram(EMG), we develop a pure mechanical sensor architecture for intent pattern recognition of lower-limb motion. The sensor system is composed of an accelerometer, a gyroscope mounted on the prosthetic socket, and two pressure sensors mounted under the sole. To compensate the delay in the control of prosthesis, the signals in the stance phase are used to predict the terrain and speed in the swing phase. Specifically, the intent pattern recognizer utilizes intraclass correlation coefficient(ICC) according to the Cartesian product of walking speed and terrain. Moreover, the sensor data are fused via DempsterShafer's theory. And hidden Markov model(HMM) is used to recognize the realtime motion state with the reference of the prior step. The proposed method can infer the prosthesis user's intent of walking on different terrain, which includes level ground,stair ascent, stair descent, up and down ramp. The experiments demonstrate that the intent pattern recognizer is capable of identifying five typical terrain-modes with the rate of 95.8%. The outcome of this investigation is expected to substantially improve the control performance of powered above-knee prosthesis.

铆接铝合金板铆钉失效缺陷检测方法研究

针对车身用铝合金板内部铆钉缺陷特征提取难度大、缺陷类型与程度识别准确率低的问题,提出一种基于高斯卷积深度信念网络与双向长短期记忆网络相结合的铆钉失效缺陷诊断模型与检测方法。首先,面向5种铆钉断裂缺陷设计试件并搭建自动检测系统,通过规划和调整探头姿态有效地降低提离效应对检测信号的影响。其次,设计双网络融合诊断模型提取和学习多维度缺陷特征信息,解决检测曲线中由时序变化特性和空间分布状态表征的缺陷信息提取难题。实验结果表明,与传统卷积网络及单一深度信念网络相比,优化后算法诊断模型的平均准确率为99.85%,相比提升了14.54%,且具有良好的通用性和鲁棒性,可实现铆钉内部缺陷的在线诊断。

A Distributed Compressed Sensing for Images Based on Block Measurements Data Fusion

Compressed sensing (CS) is a new technique for simultaneous data sampling and compression. In this paper, we propose a novel method called distributed compressed sensing for image using block measurements data fusion. Firstly, original image is divided into small blocks and each block is sampled independently using the same measurement operator, to obtain the smaller encoded sparser coefficients and stored measurements matrix and its vectors.? Secondly, original image is reconstructed using the block measurements fusion and recovery transform. Finally, several numerical experiments demonstrate that our method has a much lower data storage and calculation cost as well as high quality of reconstruction when compared with other existing schemes. We believe it is of great practical potentials in the network communication as well as pattern recognition domain.

RANDOM FOREST FOR INTERMEDIATE DESCRIPTOR FUSION IN SHOT BOUNDARY DETECTION

Shot boundary detection is the fundamental part in many real applications as video retrieval and so on. This paper tackles the problem of video segment obtaining in complex movie videos. Firstly, intermediate descriptor is proposed to depict the variation of both abrupt and gradual change in shot boundaries, which is formed by distance vector on Local Binary Pattern(LBP), GIST(GIST) or their fusion. Instead of just using the adjacent frames distance, intermediate descriptor keeps the distances between current frame and consecutive frames. It comprehensively characterizes local temporal structure, which is especially important for gradual change. For the excellent ability for feature fusion in random forests, it is adopted here to verify the fusion effect of intermediate descriptor on LBP and GIST. The whole experiments are designed on the subset of TRECVid 2013 INS(INstance Search) task to verify the effectiveness of proposed intermediate descriptor and the fusion ability for random forest. Compared with static and adaptive thresholds approaches, the best performance can be achieved by post-fusion of intermediate descriptor on LBP and GIST.

一种基于LC-BoVW模型的建筑垃圾物料识别算法

针对建筑垃圾物料的种类多、形貌易混淆等问题,构建了一种基于局部约束的视觉词袋(local constraint-bag of visual words,LC-BoVW)模型的建筑垃圾物料识别算法。首先,对建筑垃圾物料图像分块,分别提取局部颜色特征和局部二值模式特征;考虑到图像分块特征的局部相似特性,构建LC-BoVW模型分别对目标图像的显著特征进行统计。然后,基于信息融合思想对特征统计量进行融合,形成图像的判别性特征并输入到分类器中进行物料的精确识别。最后,利用自建的5类建筑垃圾物料图像数据集进行实验,实验结果表明,所提算法能够快速有效地实现建筑垃圾物料识别,平均识别准确率可达到97.92%。

Fusion of ERT Images Based on Dempster-Shafer's Evidence Theory

In this paper, an electrical resistance tomography(ERT) imaging method is used as a classifier, and then the Dempster-Shafer's evidence theory with fuzzy clustering is integrated to improve the ERT image quality. The fuzzy clustering is applied to determining the key mass function, and dealing with the uncertain, incomplete and inconsistent measured imaging data in ERT. The proposed method was applied to images with the same investigated object under eight typical current drive patterns. Experiments were performed on a group of simulations using COMSOL Multiphysics tool and measurements with a piece of porcine lung and a pair of porcine kidneys as test materials. Compared with any single drive pattern, the proposed method can provide images with a spatial resolution of about 10% higher, while the time resolution was almost the same.

临床医学专业药理学一流课程建设与实践

面对医学教育变革及复合创新人才培养的需求,在河北大学临床医学专业获批国家级一流本科专业背景下,进行药理学一流课程的建设与实践。教学团队通过分析教学现状与痛点问题,进行教学内容的优化设计,以“思政融合、科教融合、多学科交叉融合、实践融合”四维一体进行课程内容整合;采用混合教学模式;运用多维度的评价方式进行课程的改革建设。通过成绩分析及调查问卷对教学效果进行评价,综合成绩显著提升,对教学改革创新满意度高,学生与教师能力素养显著增强,一流课程建设效果显著。

Product Image Classification Based on Fusion Features

Two key challenges raised by a product images classification system are classification precision and classification time. In some categories, classification precision of the latest techniques, in the product images classification system, is still low. In this paper, we propose a local texture descriptor termed fan refined local binary pattern, which captures more detailed information by integrating the spatial distribution into the local binary pattern feature. We compare our approach with different methods on a subset of product images on Amazon/e Bay and parts of PI100 and experimental results have demonstrated that our proposed approach is superior to the current existing methods. The highest classification precision is increased by 21% and the average classification time is reduced by 2/3.

多源知识图谱事件知识融合方法研究

以事件为中心的动态知识对事件预测等应用至关重要,但现有知识图谱主要关注以实体为中心的静态知识,难以满足需求。本文提出一种融合多源知识的高质量事件知识图谱构造方法,首先定义全局事件模式,利用标签类别从源知识图谱中提取事件知识并构造临时事件知识图,提出关系扩充规则对临时事件知识图进行扩充,改进实体对齐Attce模型,基于TransD模型对多个临时事件知识图进行联合嵌入学习,以提高实体对齐和冲突发现的效率;利用事件描述完整度计算源知识图谱可信度,发生冲突时作为判别标准进行处理。经过在真实数据集上的实验,验证了该方法的准确性和有效性。

一种恒星背景的空间目标检测方法

为解决在远距离探测阶段空间目标与恒星的光谱特性和运动相似导致难以检测的问题,研究首先提出多源标准星型模式的星图识别算法,构建标准星型模式,合并多个观测星的识别结果;再提出多帧特征融合判决方法,应用星图仿真、均值漂移聚类以及置信度判决。实验结果表明,对比传统拓扑结构的星图识别算法,多源标准星型模式对星点位置噪声和星等噪声具有更强的抗干扰能力,识别率均高于95%且时间少于2 s,多帧特征融合判决方法准确识别与恒星运动相似的空间目标。该研究提出的方法可以检测恒星背景的空间目标。

不同光照模式信息融合的苹果霉心病近红外快速识别

苹果清脆甘甜,物美价廉,在我国具有极高的经济价值,是苹果产区乡村振兴的特色支柱产业。霉心病是苹果主要的内部缺陷之一,极大影响苹果的品质,若未能及时剔除将不利于苹果品牌建设,严重影响苹果产业发展。利用近红外光谱技术获取苹果在不同摆放姿态和光照模式下的光谱,去趋势(Detrend)、基线校正(Baseline)、二阶导数(SD)方法对原始光谱进行预处理,应用支持向量机(SVM)方法建立苹果霉心病判别模型,分析不同摆放姿态下光照模式对霉心病检测影响,并探索不同光照模式光谱信息融合的苹果霉心病检测方法。结果表明,在单光照模式下,180°光照模式对正常苹果有较高的判别正确率,而135°光照模式对霉心病苹果有较好的识别正确率;对于单光照模式,在果梗向右姿态和135°光照模式下获得最优的SVM模型,其模型校正集的敏感性、特异性及正确率分别为1、0.9782和0.9863,预测集的敏感性、特异性及正确率分别为0.8888、0.9565和0.9375。对于不同光照模式的信息融合,大部分光照模式的信息融合能一定程度上提高苹果霉心病判别模型的性能;在果梗向右姿态下180°与135°光照模式信息融合中获得最优的SVM模型,其模型校正集的敏感性、特异性及正确率均为1,预测集的敏感性、特异性及正确率分别为0.8888、1和0.9687,与单光照模式的最优模型相比,模型性能有所提高。该研究提供了一种基于近红外光谱和不同光照模式信息融合的苹果霉心病无损识别的新思路,也可为其他水果的内部品质无损检测提供参考。

基于推文传播模式与跨模态特征的网络谣言检测研究

[研究目的]为了有效治理网络谣言,减少网络谣言对社会稳定带来的威胁,提出充分整合帖子的多模态信息和传播模式信息对谣言进行精准识别。[研究方法]提出融合推文传播模式信息与跨模态特征的网络谣言检测模型(PPCMRD)。在推文传播特征挖掘方面,首先通过推断潜在连接补全推文传播图,接着采用双向标签图注意力模块编码推文的多个传播模式,然后通过传播模式信息融合模块捕获模式特征间的互补信息,得到帖子的传播特征;在整合多模态特征方面,该模型将帖子的文本、图像和推文传播特征集成在一起,采用跨模态共同注意力机制捕捉不同模态信息间的互补关系,得到帖子的最终嵌入表示,判断是否是谣言。[研究结论]在两个公开数据集上的实验结果表明,PPCMRD模型能够有效地检测谣言,并优于当前的基线模型。

基于非线性尺度空间与极坐标分布熵的GIS局放特征提取方法

气体绝缘组合电器(gas insulated switchgear,GIS)局部放电(partial discharge,PD)缺陷类型的准确识别对电力设备的状态评估与故障诊断至关重要,为了解决PD模式识别中特征提取准确性不足的问题,该文提出一种基于极坐标分布熵优化的非线性尺度空间特征(KAZE)提取方法。首先,对光、电单参量图谱进行非下采样轮廓波变换(non-subsampled contourlet transform,NSCT),得到包含信息更丰富的光电融合图谱;然后,利用KAZE算法提取图谱典型特征点,并依据相位、幅值、尺度信息将特征点发散至极坐标,提取子区域分布熵构成特征向量;最后,将特征信息带入经自适应增强算法(Adaboost)优化的长短期记忆网络(long short-term memory,LSTM)进行模式识别验证,并与KAZE法、统计参数法、卷积神经网络(convolutional neural networks,CNN)进行对比。结果表明,该文提出的特征提取方法在不同训练集分布下均可达到较高的识别率,最高可达91%,相较于统计参数法、CNN分别提升8.8%和4.4%,可以为提高GIS局部放电特征提取准确性提供参考。

基于多传感器融合的电力变压器内部放电定位与辨识技术

针对电力设备状态监测中传感器信息融合的难题,应用油中溶解气体、特高频、光纤超声等监测传感器构建了适用于变压器的融合传感阵列,提出了变压器内部放电的辨识策略。采用超声信号到达时间差估算实现了放电源定位,建立了多传感器监测特征融合算法,并结合K邻近(K nearest neighbors,KNN)算法实现了放电模式辨识。实验结果表明:放电源定位平均误差为45.1 mm,放电类型辨识准确率达到90.6%;所构建的多传感器融合感知策略与系统,可实现变压器内部放电源的准确定位与辨识,具有一定的应用推广价值。

基于扩散模型的印花图案生成方法设计

印花图案的设计是服装生产领域非常关键的一环,但目前人工设计的印花图案存在内容相似化、设计效率低下的问题,因此设计并实现了一种基于扩散模型的印花图案生成方法;采用深度学习技术提取、扩充现有印花图案数据集,并从颜色和类别维度生成印花图案的文本描述,完成印花图案数据集制作;使用已制作的数据集微调扩散模型,并对图像特征空间进行平铺处理,使得生成的印花图案在边界过渡处满足纺织行业四方连续的要求。对局部扩散特性进行了分析,结合文本引导,实现细节可变的图文生图效果;实验结果表明,所设计的印花图案生成方法具备生成高质量印花图案的能力,并且其特征空间平铺方法使得印花图案边界过渡处较为自然。

基于SDP信息融合的用电特征分析及负荷识别方法研究

针对多标签负荷识别信息缺失的问题,提出了基于对称点图案分解法(symmetrized dot pattern,SDP)信息融合的用电负荷特征分析及智能识别方法。针对模态混叠和残余辅助噪声问题,使用互补集合经验模态分解(complementary ensemble mode decomposition,CEEMD)分解提取电流的周期信号,提高了信号分解的鲁棒性并减小了重构误差;针对特征提取的信息缺失问题提出了基于SDP的负荷融合特性分析方法,提高了特征信息的完备性。在此基础上,提出SDP-YOLOv5的负荷识别方法,建立了SDP-YOLOv5的负荷智能识别模型。通过实验研究显示,该方法的负荷识别精度达到了98%,保证了非侵入式负荷监测水平。

高压电缆局部放电检测及识别系统研究

为了提高电气设备安全性,保证电力系统稳定运行,有必要对高压电缆局部放电进行有效检测与识别。本研究开发一套基于异源传感器数据融合的高压电缆局部放电检测与识别系统。该系统利用电场传感器检测电缆电场分布情况,发现电缆接头绝缘隐患区域,利用压力波传感器检测电缆隐患区域是否出现局部放电及其程度,利用改进的自适应阈值离散小波变换进行信号去噪处理,利用改进的格拉姆角场特征变换增强数据分类特征,利用高效信道注意力机制改进的残差卷积神经网络进行局部放电模式识别。本研究以电缆尖刺放电、气隙放电及沿面放电为对象开展测试,结果发现该系统能准确检测电缆局部放电特性,能够有效识别高压电缆缺陷的放电模式,实验室的局部放电检测率达100%,放电模式识别率达96.0%,且在工程应用环境表现优异,这对电缆使用安全及电网运行稳定具有重要意义。