Books:

  • Data Structures and Algorithms
    (in Persian)
    Mohsen Ebrahimi Moghadam,
    Azarakhsh Keipour
    , Amir Hossein Abdi
    Nasir Publishing - 1st ed. - 462 pgs. - ISBN 978-600-6927-5

    May 2013

    More Information
    This book is mostly intended for undergraduate level Data Structures course in Iranian universities. The book is written in 11 chapters. It is very useful for nationwide master's and PhD degree university entrance examinations and National Computer Engineering Olympiad. We tried hard to give the reader deep understanding about how to use basic data structures and how to design new data structures for their purposes.
    We started writing the book in 2010. AFAIK, this book was the main textbook for "Data Structures" course at Shahid Beheshti University since its publication. It is also a reference for "Data Structures and Algorithms" course at Sharif University of Technology and some other universities.
    Download link removed due to publisher's request!

Papers:

  • Real-time elliptic target detection for robotics applications

    Azarakhsh Keipour
    , Guilherme A. S. Pereira, Sebastian Scherer
    In preparation - To be submitted to 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2017)

    Vancouver, BC, Canada

    24-28th Sep 2017

  • Omnifont Persian OCR system using primitives

    Azarakhsh Keipour
    , Mohammad Eshghi, Sina Mohammadzadeh Ghadikolaei, Negin Mohammadi, Shahab Ensafi
    2013 IEEE International Conference on Industrial Technology (ICIT 2013) - Accepted, but not presented

    Cape Town, South Africa

    25-27th February 2013

    Abstract
    In this paper a model-based omnifont Persian OCR system is introduced. The system uses a set of 8 primitive elements as structural features for recognition. At first, the input scanned document is preprocessed. After normalizing the preprocessed image, text rows and sub-words are separated and then thinned. After recognition of dots in sub-words, strokes are extracted and primitive elements of each sub-word are recognized using the strokes. Finally, the primitives are compared with predefined character identification vectors in order to identify sub-word characters. The separation and recognition steps of the introduced system are done concurrently, which helps to eliminate unavoidable errors which occur during independent separation of letters. The system has been tested on documents with 14 standard Persian fonts in 6 sizes. The achieved precision is 97.06%.
    [download draft] 
  • A novel control technique for Power Factor Correction in SEPIC converter utilizing input/output voltage waveforms sampling

    Azarakhsh Keipour
    , Zeinab Sudi, Erfan Mohagheghi, Abed Hadipour Lakmehsari, Amin Hajihosseinlu
    Proceedings of
    2012 IEEE International Conference on Power and Energy (PECON 2012)

    Pages: 275 - 280
    Sutera Harbour, Sabah, Malaysia

    2-5th December 2012

    Abstract
    In this study, a simple and novel control technique for Power Factor Correction (PFC) is introduced, which is a SEPIC-based (Single-ended Primary-inductor Converter based) rectifier with high input power factor. Operation of the proposed method is based on sampling of input and output voltages in SEPIC rectifier. Using samples of input and output voltage, this control system makes the input current to follow a sinusoidal path. As a result, the line current's Total Harmonic Distortion (THD) is reduced and the input power factor is improved. This paper also compares the new method with the conventional control method for the power factor correction. In order to examine the proposed control technique comprehensively, computer simulations and experimental implementations were utilized to verify the feasibility and effectiveness of this novel control technique.
    [download] 
  • Analyzing and modeling of a new resonance inverter for low power vehicular application

    Erfan Mohagheghi,
    Azarakhsh Keipour
    , Zeinab Sudi, Mohammad Moallemi, Amin Hajihosseinlu
    Proceedings of
    2012 IEEE International Conference on Power and Energy (PECON 2012)

    Pages: 558 - 563
    Sutera Harbour, Sabah, Malaysia

    2-5th December 2012

    Abstract
    In this paper we proposed a novel inverter to convert a DC voltage to a desired AC voltage. This inverter is designed for variable inductive loads and low output power applications. We proposed a novel simple PWM method which enables the inverter to keep the output current at six times more than the rated output current, without reducing output voltage value. In addition, based on a simple use of resonance in the circuit, the new inverter can raise the output voltage to extremely high amplitude for a relatively short time. These characteristics make the proposed inverter is useful for some industrial applications such as electrical vehicle. The new configuration of the circuit consists of a unidirectional switch, two inductors (to transfer energy), a fast diode and a capacitor. With very few elements, this system changes the input DC voltage to a desired sinusoidal AC voltage. Some advantages of this system are: number of power electronics devices is low - it only uses a simple switch; it does not need dead time, in spite of conventional inverters which need a dead time between switches to prevent from short-circuiting; volume of the circuit is very small; the total harmonic distortion (THD) is greatly reduced; and it works with high efficiency. Another advantage of this inverter is capability of the circuit in boosting or bucking the input voltage to a desirable output voltage without using any DC-DC converters. It is flexible in keeping output voltage constant when the output current is increased, and it can produce the extremely higher output voltage than input voltage. The main drawback of the method, compared with conventional voltage source inverters, is that it uses more energy storage elements in spite of its low volume of power. We used simulations to prove all these statements.
    [download]  
  • Recognition of separate and joined Persian letters using primitive elements forming letters

    Shahab Ensafi, Mohammad Miremadi, Mohammad Eshghi, Mahsa Naseri,
    Azarakhsh Keipour
    Proceedings of
    17th Iranian Conference on Electrical Engineering (ICEE 2009)

    Tehran, Iran

    12-14th May 2009

    Abstract translation
    In this paper a novel algorithm for recognition of Persian letters and words using decomposition into forming primitive elements is presented. All Persian letters are constructed using 9 primitive elements. At first, using modified Hough transform, primitive elements are extracted and used to construct element array. Then, this elements are compared with Character Identification Vectors from beginning and end of array and characters are recognized. In this method, the two phases of separation and recognition are done at the same time. The algorithm does not depend on font type and size. This algorithm is implemented on 1,2,3-letter Persian subwords with 86.93% precision.
    [download] 

Theses:

  • Survey of Different Heuristics for Set-Cover Problem

    Azarakhsh Keipour

    Computer Engineering (Software) BSc. Thesis

    Department of Computer Engineering

    School of Electrical and Computer Engineering

    Shahid Beheshti University

    August 2011

    Abstract
    Set covering problem is a classical problem in Computer Science and Complexity Theory. It has many applications, such as airline crew scheduling problem, facilities location problem, vehicle routing, assignment problem, etc. Set covering problem is NP-Hard, i.e. one can not always find the exact solution in polynomial time. In this project, after precisely defining the set covering problem and needed preliminaries, we survey different efficient heuristics and approximation algorithms for this problem.
    [download] 
  • Improving the quality of Persian and Arabic Optical Character Recognition using primitive elements forming letters

    Azarakhsh Keipour

    Electrical Engineering (Electronics) BSc. Thesis

    Department of Electrical Engineering

    School of Electrical and Computer Engineering

    Shahid Beheshti University

    March 2012

    Abstract
    Optical Character Recognition (OCR) is one of the major subjects in Digital Signal Processing and Computer Vision. In this thesis we improve the quality and precision of one of the best algorithms for recognition of Persian and Arabic characters. The algorithm works by decomposing words to their containing primitive elements. The precision of this algorithm was 93/43% for 5 Persian fonts in 3 sizes of 18, 20 and 22. The precision of the algorithm after improvements is 96/48% for 14 Persian fonts in 5 sizes of 14, 16, 18, 20 and 22.
    [download] 

Reports:

  • Automated high-speed traffic monitoring and violation detection using RFID technology

    Azarakhsh Keipour
    , Mohammad Reza Mousaei, Edwin Babaians
    Tehran, Iran

    10-14th May 2015

  • Design and Implementation of a Web Software System for ACM Lab Management, and Organization of Programming Contests

    Azarakhsh Keipour

    Computer Engineering (Software) BSc. Training Report

    Department of Computer Engineering

    School of Electrical and Computer Engineering

    Shahid Beheshti University

    December 2010

    More Information
    As my Computer Engineering training course, I designed a Web Software System for management of ACM lab and organization of programming contests in summer 2010. Major parts of the system were implemented.
    [download] 
  • Construction of Process Control Boards using PLC

    Azarakhsh Keipour

    Electrical Engineering (Electronics) BSc. Training Report

    Department of Electrical Engineering

    School of Electrical and Computer Engineering

    Shahid Beheshti University

    November 2010

    More Information
    As my Electrical Engineering training course, I worked in Phase Kar Eng. Co. as an intern in summer 2010. I learned construction of Process Control Boards using Siemens Programmable Logic Controllers (PLCs).
    [download] 

Team Description Papers (TDPs):

  • 4th International AUT Robotics Competitions Service Delivery Robots League

    Azarakhsh Keipour
    , Kourosh Sartipi, Sahand Sharifzadeh, Hossein Bakhshi Golestani, Faraz Foroughi
    Sharif University of Technology (SUT) Team Description Paper

    November 2013

    More Information
    This paper provides an overview of the hardware and software of our service robot and presents the scientific achievements embodied in our AUTCup Service Delivery Robot entry. This robot has been built by the team SUT which competes on behalf of Sharif University of Technology. RobX is the champion of 2012 AUTCup and 4th National Khwarizmi Robotic Competition. It defines a Human Robot Interaction interface based on Speech, Face, Object and Gesture Recognition.
    [download] 
  • 4th Khwarizmi Robotics Competitions Service Robots League

    Azarakhsh Keipour
    , Sahand Sharifzadeh, Edwin Babaians, Farid Taherpour, Amin Malekpour, Hossein Bakhshi Golestani
    Sharif University of Technology (SUT) Team Description Paper

    November 2012

    More Information
    This Paper provides an overview of RobX robot hardware and software, and presents the scientific achievements embodied in our AUTCup/Khwarizmi Service Robots entry. This robot was built by team SUT, which competed on behalf of Sharif University of Technology in 4th National Khwarizmi Robotics Competitions. RobX defines a Human-Robot Interaction interface based on Persian Speech, Face, Object and Gesture Recognition.
    [download] 
Filename:
ICIT13.pdf
Language:
English
Pages:
5
Size:
390 kB
Filename:
PECON12Wav.pdf
Language:
English
Pages:
6
Size:
657 kB
Filename:
PECON12Veh.pdf
Language:
English
Pages:
6
Size:
435 kB
Filename:
ICEE09.pdf
Language:
Persian
Pages:
6
Size:
214 kB
Filename:
RICT05.pdf
Language:
Persian
Pages:
9
Size:
269 kB
Filename:
CEThesis11.pdf
Language:
Persian
Pages:
57
Size:
697 kB
Filename:
EEThesis12.pdf
Language:
Persian
Pages:
121
Size:
2.85 MB
Filename:
CEReport10.pdf
Language:
Persian
Pages:
56
Size:
2.37 MB
Filename:
EEReport10.pdf
Language:
Persian
Pages:
53
Size:
2.18 MB
Filename:
SUT-Kh12.pdf
Language:
English
Pages:
8
Size:
454 kB
Filename:
SUT-AUT13.pdf
Language:
English
Pages:
10
Size:
446 kB
May 2013
- The cover of my first published book (Data Structures and Algorithms).