عرض: 25 50 75 100 النتائج

نتائج البحث: 5 من أصل 5

تمييز الحروف العربية المعزولة المكتوبة بخط اليد باستخدام الشبكات العصبية == Recognition Of Isolated Handwritten Arabic Letters Using Neural Networks

اسم المؤلف: نهلة ظاهر حبيب بهية
اسم المشرف: منذر نعمان التكريتي
السنة: 2004
الموضوع الدقيق: هندسة الكهرباء
الدرجة: ماجستير
اللغة: الانكليزية
مكان الجامعة: بغداد
الصفحات الاولى:
المستخلص: يتناول هذا البحث مشكلة تمييز الحروف العربية المعزولة المكتوبة بخط اليد باستخدام الشبكات العصبية. ان نظام تمييز الحروف العربية المقترح يتكون من ثلاث مراحل وهي : مرحلة مسح صور الحروف العربية ثم مرحلة معالجة الصور واخيرا مرحلة التدريب والتصنيف.ان المرحلة الاخ | This thesis presents an algorithm for recognition of an off - line isolated handwritten Arabic letters using neural networks.The handwritten recognition system present in this work consists of three modules : - 1 - Arabic letter scanning module 2 - Arabic letter preprocessing module 3 - Learning and recognition module.The proposed neural network in the third module is trained in two stages : The first stage employs Self - Organizing Map learning algorithm for clustering the input pattern, which are based on a database of 196 letters collected from 7 independent persons.In the second stage each of the similar output cluster is considered as a subnet introduced to neural network trained by Back - propagation learning algorithm for classification. Several networks architecture are designed in the second stage using 1 - hidden and 2 - hidden layers with 5 and 28 output nodes in the output layer with different number of hidden nodes and learning rate.To examine the efficiency of the system a database of 196 letters collected from other 7 independent persons are used, (in order to test the ability of the trained network to generalize). The results show that clustering the input pattern, using 1 - hidden layer and 28 nodes in output layer improve the network performance. The system was implemented using (IBM - PC) of type Pentium 3.The programming language used to design the system was c++ version 5.02.

Design And Fpga Implementation Of Neural Network

اسم المؤلف: مثنى حاجم حمد العامري
الموضوع العام: هندسة السيطرة والنظم
السنة: 2004
الموضوع الدقيق: هندسة السيطرة والنظم
الدرجة: دكتوراه
اللغة: الانكليزية
مكان الجامعة: بغداد
الصفحات الاولى:
المستخلص: The use Artificial Neural Networks (ANN) can be a form of Artificial Intelligence (AI). The feedforward neural network has a wide application area such as pattern recognition, image compression, and classification problem. Two models of a feedforward neural network are proposed and implemented using the schematic editor of the Xilinx foundation series 2.1i. Model - 1 consists of two layers and specializes in solving linear problems. Depending on the type of application, the input layer can receive 2 to 126 input values ordered in 256x16bits RAMs. The connection weights are distributed over four 256x16bits RAMs where, the four RAMs exchange their active role in swapping operation. Model - 2 is a modified copy from Model - 1 and consists of three layers and it is responsible for classifying non - linear problems.The mathematical model of the data set (weights and inputs) is presented in a matrix multiplication format. Principle Component Analysis (PCA) is a modern method used to reduce patterns set dimensionality and hence speeds up the training phase iterations. Speeds up the training phase will eventually minimize the over all system execution time. Each model is designed and implemented in five stages without using the finite state machine. It controls the processes of the forward propagation phase, error calculation, and training algorithm. These processes are managed by many control circuits like, J - K synchronized circuit, sign - detector/sum - sub control circuit, and timers that takes the role of finite state machine. These five stages make the design easily to implemented and modified. Modification in the system parameters (No. of inputs, No. of outputs, or No. of layers) can be performed in the appropriate stage without reservation.The flexibility, low costly, and real - time operation are the main features of the proposed design. Model - 1 execution time is 2.935µs and model - 2 execution time is 2.96µs, while the costs of two models are 1927 and 2017 CLBs respectively.These features compare extremely well with other existing designs with good advantages.

الدالة الغامضة للاشارة الرادارية المرمزة == Ambiguity Function Of Radar Coded Signals

اسم المؤلف: اياد عطيه عبد الكافي الجبوري
اسم المشرف: نزار خليل وفي | وليد خالد
السنة: 2004
الدرجة: ماجستير
اللغة: الانكليزية
مكان الجامعة: بغداد
الصفحات الاولى:
المستخلص: تم دراسة انواع متعددة من تر ميزات الاشارة الرادارية حيث تم التركيز في هذه الدراسة على خواص الارتباط والغموض للتر ميزات. كذلك تم التركيز على تاثير دوبلر على تلك الخواص. تم دراسة تاثير الخطا الطوري والتعيير(Weighting) على خواص تلك الترميزات. الدراسة ال | Various types of coded radar signal are studied in this study in terms of their ambiguity and correlation properties. Doppler effect on the ambiguity and the correlation functions of these codes is studied. The effect of phase error and weighting are presented.The study related to Doppler effect covers the frequency, binary and polyphase codes in terms of the peak signal to sidelobes ratio (PSR).Linear FM, nonlinear FM, binary and polyphase codes with various windows, are dealt with respect to the sidelobes response. The simulator built using Matlab version 6. The simulation consists of a code generator; modulator; white noise adder; received signal with time and frequency shifted; matching filtering; a module to find the Doppler effect on the autocorrealtion function and a module to find the peak signal to sidelobes ratio (PSR). It is found that the m - sequences are the best binary codes in terms of PSR and the P4 code is the best Polyphase code in terms of PSR. Also, it is found that the P4 code provides better Doppler tolerant than the other codes. The m - sequences are very sensitive to Doppler shifts. Concerning the frequency codes : it is found that the nonlinear FM is more sensitive than the linear FM i.e., linear FM is Doppler tolerant while the nonlinear is not. Finally, it is found that the use of time weighting produces a response with extremely low sidelobes

ضغط الكلام باستخدام ترميز التنبؤ الخطي في مجال تحويل المويجة == Speech Compression Using Linear Prediction Coding In Wavelet Domain

اسم المؤلف: مصعب تحسين صلاح الدين
اسم المشرف: ماهر خضير محمود العزاوي
السنة: 2004
الموضوع الدقيق: هندسة الكهرباء
الدرجة: ماجستير
اللغة: الانكليزية
مكان الجامعة: بغداد
الصفحات الاولى: