Dissertation > Excellent graduate degree dissertation topics show

Optical Parallel Computing and Its Application in Image Processing and Computational Molecular Biology

Author: XuXiaoHua
Tutor: Chen
School: Yangzhou University
Course: Applied Computer Technology
Keywords: Parallel computing Molecular Biology Parallel algorithms Processor Image processing Sequence Time complexity Distance transform Sorting algorithm Fast algorithm
CLC: TP338
Type: Master's thesis
Year: 2005
Downloads: 119
Quote: 0
Read: Download Dissertation

Abstract


Light the parallel calculation accuracy of the delay, the high transmission speed, high-bandwidth, high reliability, capable of simultaneously processing a large number of data. This paper studies the optical parallel computing applications in the digital image signal processing and computational molecular biology. We introduced a common LARPBS light parallel computing model. , And in some basic operations the LARPBS on the basis of matrix multiplication and sorting algorithms; LARPBS on fast parallel image processing Hough transform and Euclidean distance transform algorithm; computational molecular biology, we LARPBS algorithm for solving the longest common subsequence and sequence alignment problems. In this article we give m · n 2 processor, constant time parallel Hough transform algorithm, the improved Hough transform algorithm to calculate the speed and efficiency of parallel algorithms, so that they are to achieve the most excellent. Euclidean distance transform (EDT), three of our design algorithm to obtain the following results: The first algorithm uses the n 2 a processor, in O (log n · loglogn completed time / logloglogn); second algorithm uses O (n 2 / loglogn) processors, the time complexity is O (logn loglogn); The third is parameterized algorithm , the use of n · d (n) · c (n) processors in O (n log n / (c (n) · logd (n))) to complete, where c (n) and d (n) satisfies 1 ≤ C (n) ≤ n and 1 lt; d (n) ≤ n In particular, if we take c (n) = n, d (n) = n epsilon , Hao ε meet O lt; epsilon ≤ 1 constant, the number of processors the algorithm n 2 epsilon , the time complexity is O (1) constant time EDT algorithm efficiency is the best algorithm. For the longest common subsequence problem, we have designed a station processor p (1 ≤ p ≤ max {m, n}), (b) (mn / p) time, and each processor using the D ( (mn) / p) space space-time consideration of optimal parallel algorithm; given a time complexity of O (logm) fast algorithm, to our knowledge, this is the fastest the longest common subsequence problem parallel algorithms. Finally, on the issue of sequence comparison, through in-depth theoretical derivation of draw sequence alignment line recurrence theorem, makes sequence comparison recurrence formula can be progressive parallel computing. Based on the line Recursive theorem, we give p (1 ≤ p ≤ n) processors, each processor requires only O ((mn) / p) space, in time O (mn / p) completed within a scalable put fast algorithm, as far as we know this is the first scalable put the time and space consideration parallel algorithm to achieve the optimal sequence comparison. The LARPBS algorithm based can prove optical parallel computing power of parallel processing capabilities and high computational efficiency, especially in light parallel computing applied to deal with the problem of computational molecular biology and will certainly become a prospect is very broad field of study.

Related Dissertations

  1. The Calculation of Heimitian Toeplitz Matrix-vector Product,O241.6
  2. Compact integral operator eigenvalue problem of numerical algorithms several studies,O177.6
  3. The Research of Fast Algorithm Based on H.264 Video Coding,TN919.81
  4. Study of Fast Motion Estimation in Video Coding,TN919.81
  5. Algorithms of Warped Discrete Fourier Transform and Speech Signal Processings,TN912.3
  6. Fast DCT new algorithms and filter structure study wavelet transform domain image noise reduction research,TN713
  7. Construction of Feature Invariants and Its Application in Object Recognition,TP391.4
  8. Image Compression Based on Lifting technology research,TN919.8
  9. Design and Application of Non-redundant Pyramid and Research on Fast Algorithm of 2-D Filter,TP391.41
  10. Research on Data Reduction Based on Extending of Rough Set,TP18
  11. Wideband Active Signal Detection Based on Continuous Wavelet Transform,TN911.7
  12. The Applied Research of PAR Method in Numerical Methods,TP311.52
  13. Research of Image Merging Method Based on Wavelet Transform,TP75
  14. Fasta Lgorithm of Pseudo-Cholesky Factorization and the Backward Error Analysis for Eigenvalue Problems,O241.6
  15. H.264 intra prediction and inter prediction of,TN919.81
  16. Narrowband communication signal source estimation precision timing differences,TN911.23
  17. H.264/AVC intra coding and decoding hardware implementation of optimization algorithms,TN919.81
  18. Real-time three-dimensional information acquisition system image processing algorithms on FPGA,TP391.41
  19. H.264 Video Coding Standard Fast Algorithm and Optimization,TN919.81
  20. SAR speckle amplitude Variational PDE Models and Algorithms,TP391.41

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Electronic digital computer (not a continuous role in computer ) > A variety of electronic digital computer
© 2012 www.DissertationTopic.Net  Mobile