检索结果-南通市图书馆

Multi-Stage-Based Siamese Neural Network for Seal Image Recognition

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computer Modeling in Engineering & Sciences 2025年第1期142卷 405-423页

作者： Jianfeng Lu Xiangye h.ang Caijin Li Renlin Xin Sh.nqing Zh.ng mahmoud Emam Sch.ol of Computer Science and Tech.ology Hangzhou Dianzi UniversityHangzhou310018China Sh.ngyu Institute of Science and Engineering Hangzhou Dianzi UniversityShaoxing312300China Faculty of Artificial Intelligence Menoufia UniversityShebin El-Koom32511Egypt

Seal auth.ntication is an important task for verifying th. auth.nticity of stamped seals used in various domains to protect legal documents from tampering and *** seal inspection is commonly audited manually to ensure document ***,manual assessment of seal images is tedious and laborintensive due to h.man errors,inconsistent placement,and completeness of th. *** image recognition systems are inadequate enough.to identify seal types accurately,necessitating a neural network-based meth.d for seal image ***,neural network-based classification algorith.s,such.as Residual Networks(ResNet)andVisualGeometryGroup with.16 layers(VGG16)yield suboptimal recognition rates on stamp ***,th. fixed training data categories make h.ndling new categories to be a ch.llenging *** paper proposes amulti-stage seal recognition algorith.based on Siamese network to overcome th.se ***,th. seal image is pre-processed by applying an image rotation correction module based on h.stogram of Oriented Gradients(h.G).Secondly,th. similarity between input seal image pairs is measured by utilizing a similarity comparison module based on th. Siamese ***,we compare th. results with.th. pre-stored standard seal template images in th. database to obtain th. seal *** evaluate th. performance of th. proposed meth.d,we furth.r create a new seal image dataset th.t contains two subsets with.210,000 valid labeled pairs in *** proposed work h.s a practical significance in industries wh.re automatic seal auth.ntication is essential as in legal,financial,and governmental sectors,wh.re automatic seal recognition can enh.nce document security and streamline validation ***,th. experimental results sh.w th.t th. proposed multi-stage meth.d for seal image recognition outperforms state-of-th.-art meth.ds on th. two establish.d datasets.

关键词： Seal recognition seal auth.ntication document tampering siamese network spatial transformer network similarity comparison network

Planting design for urban parks： Space syntax as a landscape design assessment tool

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Frontiers of Arch.tectural Research 2015年第1期4卷 35-45页

作者： Ayman h.ssaan mahmoud Reh.m h. Omar Department of Arch.tecture Faculty of Engineering Cairo University Giza 12211 Egypt Th. Canadian International College Egypt

Trees are a major factor in defining th. spatial qualities of outdoor spaces. Th.s study investigates th. influence of spatial configuration produced by th. proposed tree planting design on th. visual fields of an urban park using space syntax th.ory. Space syntax th.ory assumes th.t an urban area can be represented as a matrix of connected spaces. Th. quantitative properties of th.s matrix in th. form of syntactic measures can be measured using computer simulations. Th.s research.investigates h.w space syntax tech.iques can h.lp assess th. effect of tree configurations on th. social structure of a small-scale garden in an urban park. Such.tech.iques are assumed useful in predicting th. social structure of th. proposed space and in assessing design alternatives. An experimental study using th.ee different planting design proposals for an urban park is conducted. Data are analyzed using space syntax tech.iques. Results reveal a significant effect of planting configurations on syntactic measures across th. th.ee proposals. Th.s study suggests th.t space syntax tech.iques may h.ve a significant value in th. assessment of sch.matic planting design, especially at th. pedestrian movement level.

关键词： Landscape desisn Space syntax Plantins Urban parks Cairo

Novel Synth.sis of Some Ph.h.lazinone Derivatives

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Ch.nese Journal of Ch.mistry 2011年第7期29卷 1446-1450页

作者： mahmoud, mahmoud R. Abou-Elmagd, Wael S.I. Derbala, h.med A. h.kal, Moh.med h. Ch.mistry Department Faculty of Science Ain Shams University Abassia 11566 Cairo Egypt

Dimeth.l h.moph.h.late condensed with.isatin to give th. unexpected five membered lactone 3 rath.r th.n th. h.lf ester 1 and th. β-lactone 2. Treatment of compound 3 with.excess h.drazine h.drate afforded ph.h.lazinone carboh.drazide 4 wh.ch.represents a novel meth.d for th. synth.sis of ph.h.lazinone derivatives. Th. carh.h.- drazide 4 upon treatment with.carbon disulph.de afforded 1,3,4-oxadiazole derivative 5, wh.ch.reacted with.eth.l- ch.oroacetate to give th. S-alkylated product 6. Th. structure of compound 3 compared with.2 was discussed using h.perch.m professional （7） AM1 calculations, X-ray single crystal structure and complete spectral data.

关键词： ph.h.lazinone carboh.drazide oxadiazole triazolo oxadiazole AMx calculation condensation

DERIVATIVE SAMPLING EXPANSIONS FOR Th. LINEAR CANONICAL TRANSFORM:CONVERGENCE AND ERROR ANALYSIS

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Computational Math.matics 2019年第3期37卷 403-418页

作者： mahmoud h.Annaby Rash.d M.Ash.rabi Department of Math.matics Faculty of ScienceCairo University12613 GizaEgypt Department of Math.matics College of Arts and SciencesNajran UniversitySaudi Arabia

In recent decades,th. fractional Fourier transform as well as th. linear canonical transform became very efficient tools in a variety of approximation and signal processing *** are many literatures on sampling expansions of interpolation type for bandlimited functions in th. sense of th.se ***,rigorous studies on convergence or error analysis are *** is our aim in th.s paper to establish.sampling expansions of interpolation type for bandlimited functions and to investigate th.ir convergence and error *** particular,we introduce rigorous error estimates for th. truncation error and both.amplitude and jitter-time errors.

关键词： Linear canonical transform Sampling th.orems Truncation error Amplitude error Jitter-time error

Optimal Deep Transfer Learning Based Colorectal Cancer Detection and Classification Model

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2023年第2期74卷 3279-3295页

作者： mahmoud Ragab Maged Mostafa mahmoud Amer h.Asseri h.ni Ch.udh.y h.ith.m A.Yacoub Information Tech.ology Department Faculty of Computing and Information TechnologyKing Abdulaziz UniversityJeddah21589Saudi Arabia Center for Artificial Intelligence in Precision Medicines King Abdulaziz UniversityJeddah21589Saudi Arabia Department of Math.matics Faculty of ScienceAl-Azhar UniversityNaser City11884CairoEgypt Cancer Biology Unit King Fahd Medical Research CenterKing Abdulaziz UniversityJeddah21589Saudi Arabia Department of Medical Laboratory Sciences Faculty of Applied Medical SciencesKing Abdulaziz UniversityJeddah22252Saudi Arabia Department of Molecular Genetics and Enzymology Human Genetics and Genome Research InstituteNational Research CentreCairo12622Egypt Bioch.mistry Department Faculty of ScienceKing Abdulaziz UniversityJeddah21589Saudi Arabia Cell Biology Department Biotechnology Research InstituteNational Research CentreGiza12622Egypt

Colorectal carcinoma(CRC)is one such.dispersed cancer globally and also prominent one in causing cancer-based ***,path.logists execute CRC diagnosis th.ough.visible scrutinizing under th. microscope th. resected tissue samples,stained and fixed th.ough.h.ematoxylin and Eosin(h.E).Th. advancement of graph.cal processing systems h.s resulted in h.gh.potentiality for deep learning(DL)tech.iques in interpretating visual anatomy from h.gh.resolution medical *** study develops a slime mould algorith. with.deep transfer learning enabled colorectal cancer detection and classification(SMADTL-CCDC)*** presented SMADTL-CCDC tech.ique intends to appropriately recognize th. occurrence of colorectal *** accomplish.th.s,th. SMADTLCCDC model initially undergoes pre-processing to improve th. input image *** addition,a dense-EfficientNet tech.ique was employed to extract feature vectors from th. pre-processed ***,SMA with.Discrete h.pfield neural network(Dh.N)meth.d was applied for th. recognition and classification of colorectal *** utilization of SMA assists in appropriately selecting th. parameters involved in th. Dh.N approach.A wide range of experiments was implemented on bench.ark datasets to assess th. classification performance.A compreh.nsive comparative study h.gh.igh.ed th. better performance of th. SMADTL-CDC model over th. recent approach.s.

关键词： Colorectal cancer deep transfer learning slime mould algorith. h.perparameter optimization biomedical imaging

A Deep Real-Time Fire Prediction Parallel D-CNN Model on UDOO BOLT V8

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2022年第12期73卷 6237-6252页

作者： Amal h.Alh.rbi h.nan A.h.sni mahmoud Department of Computer Sciences College of Computer and Information SciencesPrincess Nourah Bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia

h.zardous incidences h.ve significant influences on h.man life,and fire is one of th. foremost causes of such.h.zard in most *** prediction and classification model from a set of fire images can decrease th. risk of losing h.man lives and *** promotion of fire emergency can be of great ***,construction of th.se prediction models is relevant and *** article proposes an operative fire prediction model th.t depends on a prediction unit embedded in th. processor UDOO BOLT V8 h.rdware to predict fires in real time.A fire image database is improved to enh.nce th. images quality prior to classify th.m as eith.r fire or *** proposed deep learning-based Very Deep Convolutional Networks Visual Geometry Group(VGG-16)model(Parallel VGG-16)is an enh.nced version of th. VGG-16 model,by incorporating parallel convolution layers and a fusion module for better *** experimental results validate th. performance of th. Parallel VGG-16 wh.ch.ach.eves an accuracy of 97%,compared to th. compared state-of-th.-art ***,we integrate th. prediction module into a UDOO BOLT V8 computer,wh.ch.precisely controlled th. fire alarm so th.t it can cautious people from fire in real *** th.s paper we propose a complete fire prediction model using a camera and th. UDOO BOLT V8 embedded *** experiments validate th. effectiveness and applicability of th. proposed fire model.

关键词： Fire prediction UDOO BOLT V8 deep learning

Deep Learning Prediction Model for NiCrAlY Diffusion Barrier Th.ckness for Tungsten Wires

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2022年第12期73卷 5755-5769页

作者： Amal h.Alh.rbi h.nan A.h.sni mahmoud Department of Computer Sciences College of Computer and Information SciencesPrincess Nourah bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia

In th. last decades,tech.ology h.s used Copper for IC interconnect and it h.s been th. best material used in th. wire ***,Copper is now sh.wing inefficiency as downscaling is getting *** research.starts to sh.w Tungsten(W)as a possible replacement,for its better downsizing *** scaling-down of interconnects dimension h.s to be augmented with.th.n diffusion *** is crucial to subdue tungsten diffusion in th. nickel-based th.rmal spray Flexicord(NiCrAlY)coating ***,diffusion barriers with.th.cknesses less th.n 4.3 nm do not to execute *** th. introduction of two dimensional layers,h.xagonal boron h.s been recommended as a substitute for Tungsten diffusion barrier layers with.th.cknesses less th.n 1.5 Nano meters(nm).Neverth.less,vacancies flaws may develop into a Tungsten dissemination path.wh.ch.is a problematic issue in th. manufacturing of diffusion *** energy layer density,of Tungsten atom diffusion via a di-vacancy in NiCrAlY,is computed by density functions *** h.s complex energy barrier wh.ch.is th.cker th.n oth.r materials such.as *** is due to th. sturdier contact and ch.rge variance of NI and Cr in ***,we utilize th. energy barriers of several vacancy constructions and produce a dataset to be employed in th. proposed 3-imensional deep learning model(3D-DNN).Our trained deep learning neural model can predict th. energy barrier of Tungsten diffusion th.ough.arbitrarily configured NiCrAlY with.accuracy greater th.n 98.4%in 5×5 *** results generate directors on selecting barriers th.ough.energy computation.

关键词： Deep learning tungsten NiCrAlY th.ckness barrier

Micro-Opto-Mech.nical Disk for Inertia Sensing

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Ph.tonic Sensors 2016年第1期6卷 78-84页

作者： Gh.da h. DUSh.Q Tadesse MULUGETA mahmoud RASRAS Department of Electrical Engineering and Computer Science iMicro Center MASDAR Institute Building A1 PO Box 54224 Masdar City United Arab Emirates

An optically enabled z-axis micro-disk inertia sensor is presented, wh.ch.consists of a disk-sh.ped proof mass integrated on top of an optical waveguide. Numerical simulations sh.w th.t th. optical power of laser beam propagating in a narrow silicon nitride （Si3N4） waveguide located under th. disk is attenuated in response to th. vertical movement of th. micro-disk. Th. h.gh.leakage power of th. TM mode can effectively be used to detect a dynamic range of 1 g - 10g （g=9.8 rn/s ）. At lest, th. waveguide is kept at a nominal gap of 1 μm from th. proof mass. It is adiabatically tapered to a narrow dimension of W×h.= 350×220nm2 in a region wh.re th. optical mode is intended to interact with.th. proof mass. Furth.rmore, th. bottom cladding is completely etch.d away to suspend th. waveguide and improve th. optical interaction with.th. proof mass. Th. proposed optical inertia sensor h.s a h.gh.sensitivity of 3 dB/g wh.n a 50μm-long waveguide is used （normalized sensitivity 0.5 dB/μm2） for th. vertical movement detection.

关键词： Micro-opto-Mech.nical system ph.tonic inertia sensor h.brid integration

Influence of potassium fertilization and foliar application of zinc and ph.sph.rus on growth.yield components,yield and fiber properties of Egyptian cotton(Gossypium barbadense L.)

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Plant Ecology 2008年第4期1卷 259-270页

作者： Zakaria M.Sawan mahmoud h.mahmoud Amal h.El-Guibali Cotton Research.Institute Agricultural Research CenterMinistry of Agriculture and Land Reclamation9 Gamaa StreetGiza 12619Egypt Soils Water and Environment Research InstituteAgricultural Research CenterMinistry of Agriculture and Land Reclamation9 Gamaa StreetGiza 12619Egypt

Aims Supplying optimal quantities of mineral nutrients to growing crop plants is one way to improve crop *** need to be used rationally in order to avoid a negative ecological impact and undesirable effects on th. sustainability of agricultural production *** application of nutrients also affects th. farmer’s *** order to calculate th. amount of fertilizer to be applied to crops,it is necessary to develop recommendation programmers th.t adjust nutrient rates to crop *** Experiments in two successive seasons were conducted to investigate th. effect of K fertilization and foliar application of Zn and P on yield and fiber properties of cotton *** ***(0.0 and 47.4 kg of K h.^(-1))was soil applied,wh.le ch.lated zinc(0.0 and 57.6 g of Zn h.^(-1),applied twice at 70 and 85 days after sowing‘DAS’)and ph.sph.rus(0.0,576,1152 and 1728 g of P h.^(-1),applied twice at 80 and 95 DAS)were applied to th. *** findings Dry matter yield,total ch.oroph.ll concentration,K,Zn and P uptake per plant,number of opened bolls per plant,boll weigh.,seed index,lint index,seed cotton yield per plant,seed cotton and lint yield h.^(-1) and earliness of h.rvest increased with.th. application of K,Zn and *** generally h.d no significant effect on lint percentage and fiber properties,with.exceptions,for micronaire reading and flat bundle strength.and uniformity ratio,wh.re th. mean values of th.se ch.racters were significantly increased over th. untreated control by applying K,and for th. micronaire reading in th. first season,wh.n applying P at 1728 g h.^(-1),and uniformity ratio in th. second season,wh.n applying Pat 1152 and 1728 g h.^(-1),wh.re th. mean values of th.se ch.racters were significantly increased over th. untreated control by applying *** th. conditions of th.s study,applying K fertilization at 47.4 kg h.^(-1) combined with.spraying cotton plants with.zinc at 57.6 g h.^(-1) and also with.P at

关键词： cotton fiber properties ph.sph.rus potassium yield and yield components zinc