Name:            Du Tiehua

 

Post:               Post Doctoral Research Fellow

 

 

 

 

 

 

 

 

 

 

 

 

 

Biography:

 

Du Tiehua obtained his Ph.D and M.Sc from National University of Singapore in 2007 and 2001, respectively. His research interests include: image processing, vision inspection, wavelet application and bio-image processing.

 

Project1:

 

  1. Registration of 3-dimensional image stacks showing rapidly contracting muscles

 

Three-dimensional time-lapse microscopy provides insights into the dynamic changes of cell morphology in animal development. However, movements of tissues during the acquisition of volume stacks can result in misalignments between successive optical sections. The remodeling of the muscles in Drosophila metamorphosis is an example where sporadic motions during image acquisition impede image analysis and volume visualization. Most of the existing images stack registration algorithms are aimed at the alignment of images displaying rigid objects. However, live muscles are non-rigid objects and their contractions and dilations represent non-linear transformations that cannot be properly corrected by applying purely linear registration methods. We developed a fully automatic deformable 3D image stacks registration method which is able to restore image stacks that are distorted by periodic contractions of muscles.   Experimental results show that it outperforms commercial and non-commercial image registration software.

 

 

  1. Tracking and Classification of dividing cells in 4D image datasets

 

Cell tracking is an important tool in studying the behavior of cells during the division cycles. Traditional tracking algorithms often fail to track cells successfully throughout all phases of the cell cycle, as sister chromatids split into different directions and the two daughter nuclei do not overlap with the position of the parent nucleus in the previous cycle. Cells and chromosomes also undergo dramatic changes in phenotypic appearance during mitosis. The incorporation of prior biological knowledge into the tracking algorithm will enhance our ability to automatically track cells as they progress through the various stages of the cell cycle. 

 

 

Publications:

 

  • 2-D Occluded Object Recognition Using Wavelets, The Fourth International Conference on Computer and Information Technology (CIT'04), pp. 227-232, WuHan China, 2004
  • Comparison of the Support Vector Machine and Relevant Vector Machine in Regression and Classification, International Conference on Control, Automation, Robotics and Vision, KunMing China, 2004
  • 2-D Partially Occluded Objects Recognition using Curve Moments, Seventh International Conference on Computer Graphics and Imaging, pp. 303-308, Hawaii USA, 2004
  • Bayesian Kernel Inference for 2D Objects Recognition Based on Normalized Curvature, Proceeding, 12th International Multi-Media Modeling Conference 2006, Beijing China.
  • A Wavelet Approach for Partial Occluded Object Recognition, the 1st International Symposium on Digital Manufacture(ISDM'2006), 2006,Wuhan China
  • 2-D Vision System for Detection and Measurement of Wound and Flap in Reconstructive Surgery, XVth International Conference on Mechanics in medicine and Biology (ICMMB), 2006, Singapore
  • Partial Occluded Object Recognition, International Conference on Product Design and Manufacturing Systems (PDMS), 2007, ChongQing China