4D and 5D image processing: 4D Post-processing of Gated SPECT Images
This method is based on motion-compensated temporal smoothing using a deformable content-adaptive mesh to model cardiac motion.
NO PARTIAL VOLUME EFFECT IS MODELD

Motion-Compensated 4D Processing of Gated SPECT Perfusion Studies

Here present two 4D motion-compensated reconstruction algorithms for gated cardiac SPECT.
PARTIAL VOLUME EFFECT IS MODELD

4D smoothing of gated SPECT images using a left-ventricle shape model and a deformable mesh
Similar as pervious with incorporation of heart shape model. We used clinical data.
PARTIAL VOLUME EFFECT IS MODELD
AND MIOCARDIUM SHAPE

Spatially Adaptive Temporal Smoothing for Reconstruction of Images Sequence
In this project, space-variant statistical descriptors of the temporal correlations in the image sequence data were used in order to improve image quality.

X-ray imaging methods:  

 

Mesh generation and mesh modeling of images: Content-Adaptive 3D Mesh Modeling for Representation of Voumetric Images
In this work we present a fast, content-adaptive approach for three-dimensional (3D) mesh representation of volumetric images. 
Accurate Mesh Representation of Vector-Valued (Color) Image
In this work we present a fast procedure for content-adaptive mesh representation of vector-valued (e.g., color) images.
Multy-Modality Tomographic Image Reconstruction
In this project we investigate a mesh-modeling approach for multi-modality image reconstruction. 
Content-Adaptive Mesh Modeling for Image and Tomographic Image Representation and Reconstruction
Here  we propose for the first time the use of a content-adaptive mesh model (CAMM) for tomographic image reconstruction. 
Numerical observer methodologies:  

 

 Other image processing and signal processing methods: Penalized Weighted Least Square Reconstruction for Partially Know System Response
In classical approach to tomographic image reconstruction, it is usually assumed that the system matrix is known exactly. This is an unrealistic assumption in practice. In this project we consider one approach to over come this problem by modeling the system matrix as the sum of a known part and an unknown random error that accounts for any error in the system modeling.
 

 

Hardware design projects: 
Bopak
This is a simple electronic that  counts bottles. If four bottle is counted a trigger signal is sent to filler. Filler sense the fluid level and when reached a fluid flow is stopped and a new cycle begins.  
Bopak1L-S
This electronic is designed for to control the bottle filling machine.
PLC (Programmable Logic Controller) for industrial application.
This is general purpose PLC processor card.
 

 

Interesting Class projects:
Speech Processing
Linear Prediction Analysis
Pitch Tracking, Voicing Discrimination, and LP-Vocoding Synthesis
Physiological Systems and Modeling (375kb)
Glucose and Insulin Responses Modeling by MATLAB 
Object Oriented Programming
1D and 2D signals Class
Mandelbrot fractal applet
Multidimensional Signals
Fourier Representation of Image
Motion Estimation Using Block Matching
Signal and Data Compression
Scalar and Vector Quantizer Design for Quantization of Speech Samples
Performance Comparison Between Vector and JPEG Quantization of Images