• ShapeMatching, Kamani

    Severe Weather Forecasting

    NSF Research Project on weather forecasting using radar images

    In this project we use radar images and apply a novel shape matching technique, along with machine learning algorithms, in order to detect bow echo patterns, which are associated with severe weather conditions.

    The ultimate goal is to forecast these bow echo patterns beforehand, so we can alert people about upcoming severe weather incident.

    Github Repository: and
  • CulturalHeritage Mohammad Mahdi Kamani

    Image Processing in Cultural Heritage

    using Image Processing techniques for investigating Cultural Heritage

    This project was my M.Sc. thesis under supervision of prof. Marvasti and prof. Amini at Sharif University of Technology. The exact name of thesis is: “Image Processing in paintings using Multispectral imaging.”

    Developing of imaging instruments and technology, today capturing very high spatial resolution and multiband images is feasible. One of the greatest application of these high resolution, multiband images is in Cultural Heritage. Curators with using these images can investigate their antiquities without invasion of these priceless heritage. In this project, we have multiband images from works of some of greatest painters in Europe such as Monet. The goal of this project is that we design a system which can compare differences in frequency layers automatically.

  • Fuzzy Speech Recognition Mohammad Mahdi Kamani

    Fuzzy Speech Recognition

    Using new approach in Speech Recognition by means of Fuzzy logic.

    p>This project was my B.Sc. thesis in partnership with Karen Khatami under supervision of Prof. Saeed Bagheri Shouraki a professor at the Electrical Engineering Department of Sharif University.

    The exact name of thesis was: Comparison between FEMM and HMM algorithms on Speech Recognition in noisy environments.

    Nowadays Speech Recognition is one of the most important applications in technology. One of algorithms which is widely used in this area is HMM (Hidden Markov Model). HMM is highly sensitive to noise. The Computer Engineering Department at Sharif University invented a new approach to speech recognition based on Fuzzy Logic named FEMM (Fuzzy Elastic Matching Machine). In this project we compared these two methods in noisy environments; the results revealed that FEMM is more immune to noise than HMM.

  • BI Mohammad Mahdi Kamani

    Business Intelligence

    Turning raw data into knowledge

    Living in the internet era, nowadays companies make huge amounts of data without extracting comparable information out of them. Business Intelligence is a new approach to deal with the big amount of data. Making Decisions easier, faster and more efficiently are the ultimate goals of BI.

    Tadbir Pardaz computer group is supporting lots of Brokerages, Securities and Mutual Funds in Iran’s capital market. We decided to implement BI on mutual funds for the first time in Iran and our goal is to implement BI in the whole capital market of Iran.