{"currentpage":1,"firstResult":0,"maxresult":10,"pagecode":5,"pageindex":{"endPagecode":5,"startPagecode":1},"records":[{"abstractinfo":"气水两相流在生产过程与科学研究中有十分重要的地位.流型识别作为两相流研究的基础,对于两相流过程参数精确测量起重要作用.信息融合技术可消除多传感器信息间可能存在的冗余和矛盾,降低其不确定性,实现比单一信息源对被测过程更完全、准确、可靠的测量和描述.本研究分别采用数据级融合与特征级融合实现双截面电阻层析成像系统两测量截面间测量信息融合,结合支持向量机实现水平管道气水两相流流型识别率的提高.通过识别结果的对比,研究了这两类信息融合方法提高流型识别率的原因.","authors":[{"authorName":"谭超","id":"97477c6a-78a9-4dec-ae8c-3041e15d7f8b","originalAuthorName":"谭超"},{"authorName":"董峰","id":"855e3eaf-7c11-45e6-b0a3-109f1f4785b0","originalAuthorName":"董峰"},{"authorName":"徐遥远","id":"0445a1ed-b4ae-40e4-b97b-28953cf016cf","originalAuthorName":"徐遥远"}],"doi":"","fpage":"785","id":"9fa8677c-d65c-4965-9e59-994d2c183d18","issue":"5","journal":{"abbrevTitle":"GCRWLXB","coverImgSrc":"journal/img/cover/GCRWLXB.jpg","id":"32","issnPpub":"0253-231X","publisherId":"GCRWLXB","title":"工程热物理学报 "},"keywords":[{"id":"074660b8-5151-4b21-80b3-a0a77c2c5a7d","keyword":"电阻层析成像","originalKeyword":"电阻层析成像"},{"id":"c17045c4-0dcf-4a15-8bc3-da5b9887fbcd","keyword":"流型识别","originalKeyword":"流型识别"},{"id":"cff2eb80-d712-44f1-afad-6f23fb83db04","keyword":"特征提取","originalKeyword":"特征提取"},{"id":"d4ba60eb-a574-4871-a5f8-1a5af547f554","keyword":"数据融合","originalKeyword":"数据融合"},{"id":"1c5569e2-bd3e-4287-8128-4684e27005d8","keyword":"特征融合","originalKeyword":"特征融合"}],"language":"zh","publisherId":"gcrwlxb201005017","title":"用于气水两相流流型识别的ERT信息融合方法","volume":"31","year":"2010"},{"abstractinfo":"利用单个特征识别强噪声中的弱小运动目标,常因所提取的目标特征与噪声特征易混淆而导致高的虚警率.提出一种新的基于多特征融合的弱小运动目标识别方法.分析了弱小运动目标的连续相关性、面积及质心位置偏移这三个特征的可靠性及提取方法,对获取的特征值进行归一化后采用多特征融合的方法构造更具有鲁棒性的联合特征,确定了以具有最大多特征融合值为真实目标的决策方法.通过与采用单一特征的目标识别方法进行比较,证明了提出的多特征融合方法能更准确地识别弱小运动目标.","authors":[{"authorName":"雍杨","id":"722ac52f-12ae-42c0-ba79-e249ba973927","originalAuthorName":"雍杨"},{"authorName":"王敬儒","id":"9e1629c8-498f-47a5-a74f-b5b347543d3d","originalAuthorName":"王敬儒"},{"authorName":"张启衡","id":"7f4bd670-11e2-45a7-9fc5-45c93a448efe","originalAuthorName":"张启衡"}],"doi":"10.3969/j.issn.1007-5461.2006.05.003","fpage":"594","id":"44fb5167-c930-4c3a-8080-90354b04196f","issue":"5","journal":{"abbrevTitle":"LZDZXB","coverImgSrc":"journal/img/cover/LZDZXB.jpg","id":"53","issnPpub":"1007-5461","publisherId":"LZDZXB","title":"量子电子学报 "},"keywords":[{"id":"a1f70c9f-2ed0-4b3e-8474-1102274ace36","keyword":"图像处理","originalKeyword":"图像处理"},{"id":"f3344f34-706d-49c6-b8f0-65e5f779a33d","keyword":"目标识别","originalKeyword":"目标识别"},{"id":"578b461f-4596-4e15-976f-fb3ed78ab594","keyword":"多特征融合","originalKeyword":"多特征融合"},{"id":"1a99f5d0-707f-41f2-9a40-160c30b508d8","keyword":"弱小运动目标","originalKeyword":"弱小运动目标"},{"id":"831f043a-57ad-45b2-ac7b-197942423902","keyword":"特征提取","originalKeyword":"特征提取"}],"language":"zh","publisherId":"lzdzxb200605003","title":"基于多特征融合的弱小运动目标识别","volume":"23","year":"2006"},{"abstractinfo":"为了对各类自然场景中的显著目标进行检测,本文提出了一种将图像的深度信息引入区域显著性计算的方法,用于目标检测.首先对图像进行多尺度分割得到若干区域,然后对区域多类特征学习构建回归随机森林,采用监督学习的方法赋予每个区域特征显著值,最后采用最小二乘法对多尺度的显著值融合,得到最终的显著图.实验结果表明,本文算法能较准确地定位RGBD图像库中每幅图的显著目标.","authors":[{"authorName":"杜杰","id":"7d318bae-a020-42f4-9392-dc1053dc50c4","originalAuthorName":"杜杰"},{"authorName":"吴谨","id":"00bfcbfd-8e63-4fb5-a8c7-9244ce012a0e","originalAuthorName":"吴谨"},{"authorName":"朱磊","id":"77ff616c-c1af-4579-8fe9-e7977a381c92","originalAuthorName":"朱磊"}],"doi":"10.3788/YJYXS20163101.0117","fpage":"117","id":"cb41456e-c0a1-4666-9a2b-224be32f2d9b","issue":"1","journal":{"abbrevTitle":"YJYXS","coverImgSrc":"journal/img/cover/YJYXS.jpg","id":"72","issnPpub":"1007-2780","publisherId":"YJYXS","title":"液晶与显示 "},"keywords":[{"id":"ed1575b6-4e87-4895-8482-1a503a1f5b42","keyword":"目标检测","originalKeyword":"目标检测"},{"id":"45345145-d98a-4cd9-a0be-66b86b51bcc1","keyword":"深度信息","originalKeyword":"深度信息"},{"id":"c7e95e5f-ad53-47ae-a667-63fd0b28d2f1","keyword":"区域特征","originalKeyword":"区域特征"},{"id":"50ccdf80-f79a-4d9e-b254-e769dc351dcf","keyword":"随机森林","originalKeyword":"随机森林"},{"id":"7930ab17-d7b9-4365-b372-106405e62ec8","keyword":"监督学习","originalKeyword":"监督学习"}],"language":"zh","publisherId":"yjyxs201601015","title":"基于区域特征融合的RGBD显著目标检测","volume":"31","year":"2016"},{"abstractinfo":"针对目标跟踪中的目标尺度变化、遮挡、光照变化、相似目标混淆等问题,本文提出多特征融合的协同相关跟踪算法.首先,本文用多种特征构建目标外观模型,提高目标模型的鲁棒性,增强跟踪的抗形变能力和抗光照变化能力.然后,利用定点优化策略,解决多模板滤波优化问题,获得最佳滤波参数,通过多模板相关滤波算法估计目标位置,利用改进的尺度池方法解决目标尺度变化问题.最后,利用目标置信度判别跟踪目标是否发生遮挡,当目标发生遮挡时,利用CUR滤波模块重新检测目标,解决遮挡情况下跟踪任务.本文利用OTB-2013数据集中的方法测试本文算法,实验表明本文算法的整体成功率和精确度为0.622和0.830,本文算法在目标发生尺度变化、遮挡、光照变化、相似目标混淆等问题情况下,能准确、可靠地跟踪目标,具有一定研究价值.","authors":[{"authorName":"毛宁","id":"312c4a6f-21ef-49c8-aad3-67b538c204cd","originalAuthorName":"毛宁"},{"authorName":"杨德东","id":"eb34770f-24b0-4c0c-985b-ebafe33d3a98","originalAuthorName":"杨德东"},{"authorName":"杨福才","id":"7f6a72cc-93cf-4157-939d-281fff567864","originalAuthorName":"杨福才"},{"authorName":"蔡玉柱","id":"2a510c41-3d1b-4f33-bb47-9f24840e7611","originalAuthorName":"蔡玉柱"}],"doi":"10.3788/YJYXS20173202.0153","fpage":"153","id":"132b8893-d525-40ea-990a-f62bd3fecbd4","issue":"2","journal":{"abbrevTitle":"YJYXS","coverImgSrc":"journal/img/cover/YJYXS.jpg","id":"72","issnPpub":"1007-2780","publisherId":"YJYXS","title":"液晶与显示 "},"keywords":[{"id":"42c4bf84-07a7-494f-8c9d-e82863c5db57","keyword":"机器视觉","originalKeyword":"机器视觉"},{"id":"a3602f46-1f48-4eb7-9de6-a52fa6a9614b","keyword":"目标跟踪","originalKeyword":"目标跟踪"},{"id":"eeb0c710-30e3-4fed-ab79-689d8140fa00","keyword":"多模板协同滤波","originalKeyword":"多模板协同滤波"},{"id":"1f6dd433-4609-4b74-a819-bff2ee57944b","keyword":"多特征融合","originalKeyword":"多特征融合"}],"language":"zh","publisherId":"yjyxs201702012","title":"多特征融合的多模板协同相关跟踪","volume":"32","year":"2017"},{"abstractinfo":"针对高分辨率遥感影像多尺度、空间分布复杂以及特征繁多的特点,从遥感影像特征提取的尺度效应以及各类地物显著性特征各异入手,提出了基于多尺度多特征融合的高分辨率遥感影像分类方法.该方法构建最优尺度分割函数模型,寻找出各地物的最优尺度,分别提取影像的纹理、颜色和形状特征.在此基础上利用各地物特征的显著性差异实现多尺度下多特征的加权融合.该加权融合方法突破了常规最优尺度分割算法未能充分考虑各类地物特征差异性的局限性,通过分析各类地物的显著性,建立了各个特征在分类中所占权重的模型.实验结果表明:相对传统无监督分类算法,该方法准确率提高约7％,且运行效率高.","authors":[{"authorName":"陈苏婷","id":"be876fb0-1cab-424a-a02a-cc76bc707116","originalAuthorName":"陈苏婷"},{"authorName":"王慧","id":"d5b3441d-f713-4840-b2d5-3d866fd65e93","originalAuthorName":"王慧"}],"doi":"10.3969/j.issn.1007-5461.2016.04.006","fpage":"420","id":"fdaccd4c-50a5-4b79-8d34-76d1a265fca5","issue":"4","journal":{"abbrevTitle":"LZDZXB","coverImgSrc":"journal/img/cover/LZDZXB.jpg","id":"53","issnPpub":"1007-5461","publisherId":"LZDZXB","title":"量子电子学报 "},"keywords":[{"id":"20380bd4-11a3-4057-94a4-59e2e829d6b0","keyword":"图像处理","originalKeyword":"图像处理"},{"id":"143dd679-032d-4b3d-8a7a-a24c3cf9a853","keyword":"影像分类","originalKeyword":"影像分类"},{"id":"ac485b0f-7c58-4e08-88d5-fdae8b597291","keyword":"多尺度特征融合","originalKeyword":"多尺度特征融合"},{"id":"199a125e-26e0-4951-8769-2a631d8a5a1f","keyword":"最优分割尺度函数","originalKeyword":"最优分割尺度函数"},{"id":"54fc19f6-a0c4-42aa-bcf2-ec9cadfe09f1","keyword":"显著性特征","originalKeyword":"显著性特征"}],"language":"zh","publisherId":"lzdzxb201604006","title":"多尺度多特征融合的高分辨率遥感影像分类","volume":"33","year":"2016"},{"abstractinfo":"为了提升多模态图像融合精度,提出了一种局部化非下抽样剪切波变换与脉冲耦合神经网络相结合的图像融合方法.首先,利用局部化非下抽样剪切波对源图像进行多尺度、多方向分解;然后,在分解后的各子带图像中,利用局部区域奇异值构造的局部结构信息因子作为PCNN神经元链接强度.经过脉冲耦合神经网络点火处理,获取子带图像的点火映射图,通过判决选择算子,选择各子带图像中的明显特征部分生成子带融合图像;最后,应用局部化非下抽样剪切波逆变换重构图像.选用多组不同模态的图像进行实验,并对实验结果进行了客观评价.实验结果表明,本文提出的融合方法在主观和客观评价上均优于一些典型融合方法,可获得更好的融合效果.","authors":[{"authorName":"陈广秋","id":"24cd53d9-43ba-470c-81fa-bba5c9d0d203","originalAuthorName":"陈广秋"},{"authorName":"高印寒","id":"ef3c75bf-06cf-4c75-acea-4cbfd5221f01","originalAuthorName":"高印寒"},{"authorName":"才华","id":"059b0dc8-86bd-4aa0-9e8a-51664a5f3be5","originalAuthorName":"才华"},{"authorName":"刘广文","id":"75bffdaa-253a-4d6c-9ef3-f98ca1c39c1d","originalAuthorName":"刘广文"},{"authorName":"段云鹏","id":"607d6b46-84e0-48c3-b5ad-00578657b7c5","originalAuthorName":"段云鹏"}],"doi":"10.3788/YJYXS20153004.0701","fpage":"701","id":"581ad503-a9de-453d-b772-52c4b6f8e54e","issue":"4","journal":{"abbrevTitle":"YJYXS","coverImgSrc":"journal/img/cover/YJYXS.jpg","id":"72","issnPpub":"1007-2780","publisherId":"YJYXS","title":"液晶与显示 "},"keywords":[{"id":"b4d07cb7-5848-4796-8483-40c16901868d","keyword":"图像处理","originalKeyword":"图像处理"},{"id":"e237c4a2-6541-464d-bda9-ffb318534da7","keyword":"局部化非下抽样剪切波","originalKeyword":"局部化非下抽样剪切波"},{"id":"08747f0b-e554-4678-ac98-3116550cbd27","keyword":"平移不变性","originalKeyword":"平移不变性"},{"id":"a91f1c33-d1f1-450b-8bc0-48fc4c01f8e2","keyword":"脉冲耦合神经网络","originalKeyword":"脉冲耦合神经网络"},{"id":"2dcb666e-ff7b-4764-a15c-3b44425cb83b","keyword":"链接强度","originalKeyword":"链接强度"}],"language":"zh","publisherId":"yjyxs201504025","title":"局部化NSST与PCNN相结合的图像融合","volume":"30","year":"2015"},{"abstractinfo":"为了提升红外与可见光图像融合精度,提出了一种基于局部区域奇异值分解的自适应PCNN红外与可见光图像融合算法.利用局部区域奇异值构造局部结构信息因子,作为PCNN对应神经元的链接强度.经过PCNN点火处理,获得源图像的点火映射图,通过比较选择算子,选择源图像中明显特征部分生成融合图像.采用多组红外与可见光图像进行融合实验,并对融合结果进行客观评价.实验结果表明本文提出的算法在主观和客观评价上均优于已有文献的一些典型融合算法,可获得更好的融合效果.","authors":[{"authorName":"陈广秋","id":"e82c904f-05de-4927-9651-5e7b78bba74b","originalAuthorName":"陈广秋"},{"authorName":"高印寒","id":"ccc50b4e-7253-42f0-9812-e52ce3e8da90","originalAuthorName":"高印寒"},{"authorName":"段锦","id":"dd09a46c-eebd-4b00-ab5e-6c57f2e0f123","originalAuthorName":"段锦"},{"authorName":"韩泽宇","id":"7ea6ecc4-bc91-44b5-bb98-80a2c8f1dc5c","originalAuthorName":"韩泽宇"},{"authorName":"才华","id":"3c5dfeee-dd5b-4f39-8fd5-7602b532c803","originalAuthorName":"才华"}],"doi":"10.3788/YJYXS20153001.0126","fpage":"126","id":"6fcd210a-dbcf-4c72-bd1d-786546e86d7a","issue":"1","journal":{"abbrevTitle":"YJYXS","coverImgSrc":"journal/img/cover/YJYXS.jpg","id":"72","issnPpub":"1007-2780","publisherId":"YJYXS","title":"液晶与显示 "},"keywords":[{"id":"0d714771-e6d0-4c85-8548-4e3412711148","keyword":"图像融合","originalKeyword":"图像融合"},{"id":"dea7f0f7-68fb-4ac0-ab8b-c35c54f43512","keyword":"奇异值分解","originalKeyword":"奇异值分解"},{"id":"cb719c9b-fb5e-4f03-8d67-3844d5d3d8c8","keyword":"局部结构信息因子","originalKeyword":"局部结构信息因子"},{"id":"c8315a6d-b465-4ce9-8697-0a71d557e493","keyword":"点火映射图","originalKeyword":"点火映射图"},{"id":"8b8a53fc-35f3-453a-9078-8287410342f1","keyword":"链接强度","originalKeyword":"链接强度"}],"language":"zh","publisherId":"yjyxs201501020","title":"基于奇异值分解的PCNN红外与可见光图像融合","volume":"30","year":"2015"},{"abstractinfo":"针对量子化算法对图像融合的缺点,提出量子克隆多宇宙算法.首先对量子克隆、变异和选择变换获得新的量子群,然后将宇宙各自独立化并且内部为并行拓朴结构,采用量子旋转门更新量子宇宙个体,宇宙之间联合交叉,实现信息的交流,最后给出了图像融合流程.实验结果表明,该方法能够更加有效、准确地融合图像中的特征,是一种有效可行的图像处理算法.","authors":[{"authorName":"邵明省","id":"2fdf0ede-76b1-48f9-a306-634e6df94f4d","originalAuthorName":"邵明省"},{"authorName":"杜广朝","id":"97cc79be-1e22-4ae2-b793-866bd8b283df","originalAuthorName":"杜广朝"}],"doi":"10.3788/YJYXS20122706.0837","fpage":"837","id":"1c63af80-7de1-45f7-bb4f-63c1176808cc","issue":"6","journal":{"abbrevTitle":"YJYXS","coverImgSrc":"journal/img/cover/YJYXS.jpg","id":"72","issnPpub":"1007-2780","publisherId":"YJYXS","title":"液晶与显示 "},"keywords":[{"id":"9dd4e1e6-dda3-4746-8c51-3d6d70171a68","keyword":"多宇宙","originalKeyword":"多宇宙"},{"id":"c898f4eb-57c3-4f3d-a38b-29c966864577","keyword":"并行","originalKeyword":"并行"},{"id":"07a050d4-1fe3-4dc4-aa7d-19cafcd12f24","keyword":"量子克隆","originalKeyword":"量子克隆"},{"id":"e7a1675a-8ca9-4c93-81d0-83b0954e1d1a","keyword":"图像融合","originalKeyword":"图像融合"}],"language":"zh","publisherId":"yjyxs201206021","title":"基于量子克隆多宇宙算法的图像融合研究","volume":"27","year":"2012"},{"abstractinfo":"移动粒子半隐式（Moving Particle Semi-implicit，MPS）数值方法在追踪汽液相界面上较传统网格方法有很大优势，本研究基于MPS方法对过冷水中单个蒸汽泡的冷凝行为进行了数值模拟研究。计算结果与Kamei的实验结果符合较好并表明，汽泡冷凝寿命与汽泡初始尺寸呈近似线性关系，低过冷度下大汽泡的变形会加速其冷凝，高过冷度下会出现冷凝波动现象。此外还利用MPS方法对汽泡对绝热融合行为进行了数值模拟，分析了汽泡在融合过程中的形变特性、融合前后汽泡上升速度的变化。本研究揭示了直接接触汽泡冷凝换热及汽泡对绝热融合行为的一些规律特征，也为MPS进一步应用于汽泡动力学数值模拟打下基础。","authors":[{"authorName":"陈荣华","id":"ccb3425a-69b1-46ec-9afa-781599d4eeab","originalAuthorName":"陈荣华"},{"authorName":"田文喜","id":"79bc9a70-9303-4c72-8bcc-a75a575f20c3","originalAuthorName":"田文喜"},{"authorName":"左娟莉","id":"a4066c61-e29b-42e7-9d9d-7482ad86628d","originalAuthorName":"左娟莉"},{"authorName":"苏光辉","id":"ca04f938-1153-4d05-97d4-2ddd2dbfa290","originalAuthorName":"苏光辉"},{"authorName":"秋穗正","id":"32f23c1d-603a-4242-a2dd-03f7dc48d18d","originalAuthorName":"秋穗正"}],"doi":"","fpage":"1876","id":"c6b61335-cf3c-4a82-a33f-90e8141b7370","issue":"11","journal":{"abbrevTitle":"GCRWLXB","coverImgSrc":"journal/img/cover/GCRWLXB.jpg","id":"32","issnPpub":"0253-231X","publisherId":"GCRWLXB","title":"工程热物理学报 "},"keywords":[{"id":"1dd2ea1f-9a17-4bee-b7b7-823131d54028","keyword":"移动粒子半隐式方法","originalKeyword":"移动粒子半隐式方法"},{"id":"b08e9206-50f6-4b33-b858-1a51f9e1faf9","keyword":"汽泡冷凝","originalKeyword":"汽泡冷凝"},{"id":"e2fd9c9b-ea44-4e20-a7b6-18dd2fad9841","keyword":"汽泡融合","originalKeyword":"汽泡融合"}],"language":"zh","publisherId":"gcrwlxb201111020","title":"基于MPS方法的汽泡冷凝与融合特性研究","volume":"32","year":"2011"},{"abstractinfo":"在小波包域提出了一种多传感器图像融合和双水印算法.首先,利用HIS变换和小波包变换将多光谱和全色图像分解为多个高低频子带,根据小波包域系数特点,低频部分采用基于区域平均能量加权算法的规则进行融合,高频部分采用绝对值取大的规则进行融合.然后,在高低频图像融合系数分别嵌入一个水印,低频水印利用了离散余弦变换的聚能去相关能力,高频水印利用了图像纹理子块特征.最后,对嵌入的双水印融合图像进行攻击和分析.实验结果显示,融合图像在保留多光谱图像光谱信息的基础上有效提高了空间分辨率;加水印的融合图像具有良好的不可视性和鲁棒性.","authors":[{"authorName":"李新娥","id":"ac3121aa-9ed4-4685-b3e2-072a5694d502","originalAuthorName":"李新娥"},{"authorName":"班皓","id":"f976e5ed-88fd-4cf2-ac2f-1b0e82dda0cf","originalAuthorName":"班皓"},{"authorName":"任建岳","id":"10d144c5-0cdd-406a-b9a1-b10b21d83e85","originalAuthorName":"任建岳"},{"authorName":"金龙旭","id":"c8b9e1e7-0604-43d3-9fbf-50e4b7da2203","originalAuthorName":"金龙旭"},{"authorName":"李国宁","id":"548aba72-6839-4f1e-a6a6-e6883c47b93e","originalAuthorName":"李国宁"}],"doi":"10.3788/YJYXS20142902.0286","fpage":"286","id":"a20aeb3b-a39f-4c6d-a71d-5f01cc85b3e6","issue":"2","journal":{"abbrevTitle":"YJYXS","coverImgSrc":"journal/img/cover/YJYXS.jpg","id":"72","issnPpub":"1007-2780","publisherId":"YJYXS","title":"液晶与显示 "},"keywords":[{"id":"cd4c913a-6173-4f30-828c-688dfc8a2959","keyword":"图像融合","originalKeyword":"图像融合"},{"id":"b2d7d647-4ccb-4ca0-a30c-55abfef8e1d4","keyword":"数字水印","originalKeyword":"数字水印"},{"id":"0e2cbed8-b327-43c6-987b-26c95c1dec9e","keyword":"小波包变换","originalKeyword":"小波包变换"},{"id":"9a3252e8-77b6-4b02-8b66-075409cfad07","keyword":"离散余弦变换","originalKeyword":"离散余弦变换"},{"id":"58f6437a-6268-4e68-9005-cf79e1e103a0","keyword":"HIS变换","originalKeyword":"HIS变换"}],"language":"zh","publisherId":"yjyxs201402022","title":"一种多传感器图像融合与数字水印技术","volume":"29","year":"2014"}],"totalpage":1308,"totalrecord":13080}