Multimedia Systems in Telecommunication, Medicine and Remote Sensing applications

Multimedia Systems in Telecommunication, Medicine and Remote Sensing applications

Research Assistant: Dr. Dimitrios Alexiadis
Academic Assistant: Dr. Nikolaos Mitianoudis
Courses covered:

Multimedia Content Management

Medical Imaging

GIS and Remote Sensing


Under this virtual laboratory, several interactive exercises through Graphical User Interfaces (GUIs) will be implemented, through which the students will be able to practice some of the theoretical aspects taught in class. This virtual laboratory will cover the following topics:


1. Audio-Image-Video Encoding and Compression. More specifically, the students will interact with:


• The concept of transform coding, i.e. coding in transform domain, in which the energy-information tends to be concentrated in a few major coefficients.


• The concept of Scalar and Vector Quantization, which is the essential part of “lossy” encoders.


• Entropy coding (encoding Huffman, arithmetic coding, etc.), which is the last subsystem of modern codecs, which does not result in loss of information (lossless).


• Audio Coding : the compression standard MP3 (MPEG Audio Layer III). Attention will be paid mainly on the lossy psychoacoustic masking, based on the limitations of the human hearing system.


• Still Image Coding: the compression formats JPEG and JPEG-2000 will be visualized with special emphasis on the 2-D Discrete Cosine Transform (DCT) and the 2-D Discrete Wavelet Transform (DWT).


• Compression of moving pictures: The encoding standards MPEG1-MPEG2 and MPEG-4 will be presented. In the case of MPEG1-MPEG2, apart from a single frame coding (intra-frame coding), we will be focusing on motion estimation (motion vectors) between successive frames. In the case of MPEG-4, the focus will be on the separation of moving objects, the encoding of shape and motion of which is the basic idea of the model.





2. Streaming audio and video through communication channels are susceptible to noise and errors. This study will study the impact of transmission errors and packet loss on the quality of decoded signals and the bit-error rate. Forward Error Correction / Detection techniques that are 'hiding' the error in the decoder.





3. Equalization - Noise Reduction - Registration - fusion of medical images. More specifically:


• Image Equalization: histogram equalization and local adaptive methods for image enhancement.


• Image Noise Reduction: Noise Reduction methods with simple filters (moving average, middle) and shrinkage methods using 2-D wavelet 2-D directional local bases (curvelets, contourlets) or 2D self-trained bases using Independent Component Analysis.


• Image Registration and Fusion: The concept of registration and fusion of images acquired from different modality sensors.



4. Fusion Applications on remote sensing images.


With the help of the corresponding interactive graphical display (GUI), students can experiment with the parameters and observe the effect of these on the performance of remote sensing imaging fusion algorithms (pansharpening).