OBJECTIVE
Image processing development

SUMMARY

• 4+ years experience on developing algorithm in signal and image processing, image segmentation and recognition, video compression
• 3+ years experience on developing parallel algorithm on image processing, searching algorithm, using MPI, CUDA, OpenCL

EDUCATION
Ph.D. in Electrical and Computer Engineering, Oct. 2011

• Mississippi State University, Mississippi State, Mississippi, GPA: 4.0/4.0

B.E. in Electronic Information Engineering, Jul. 2007

• Northwestern Polytechnical University, Xi’an, PRC, GPA: 86/100

EXPERIENCE
Confidential,
MS
Research Assistant, 01/2008 – Present

• Human face detection
• Applied different color spaces for segmentation
• Implemented morphological processing to reduce the little spaces between faces
• Matched filter was applied to detect the face position based on the template matching method
• Implemented in Matlab
• Image texture classification
• Implemented a “Gabor-like” filterbank and a support vector machine classifier system for image texture classification
• Studied the samples from Brodatz texture data and the algorithm was proven effective on accuracy and kappa coefficient
• Implemented in Matlab
• Build JPEG-like encoder and decoder for video compression
• Implemented the “JPEG-like” encoder and decoder system composed of the DCT, quantization, Huffman coding and its corresponding decoder. High level aspects of standards, such as headers and synchronization markers were ignored
• Studied the distortion-rate curves for the system compared with arithmetic-encoded DPCM system
• Implemented in C
• Parallel computing in band selection of hyperspectral image
• Developed the parallel computing in supervised band selection of hyperspectral image by using cluster and GPU
• Studied the search algorithm speedup and performance of both parallel computing modes
• Implemented the project in MPI, CUDA and C
• Parallel optimization-based spectral transformation for detection and classification of buried radioactive materials
• Developed the cluster and GPU parallel algorithm in the optimization of channel combination in the spectral transformation
• Implemented the parallel computing of PSO evolution and searching process
• Implemented in MPI, CUDA and C
• Parallel search for intelligent puzzles
• Developed the parallel cluster algorithm for the intelligent puzzles search problems by using master and slave parallel scheme
• Studied the effect of latency time on total running time when the number of tasks sent to each client is varied
• Applied MPI and C in this project
• Noise-adjusted principal component analysis for buried radioactive target detection and classification
• Proposed several spectrum transformations to reduce the negative effect of signal noise
• Utilized principle component analysis (PCA) and noise adjusted PCA to reduce dimensionality and extract feature
• Modeled by Matlab and implemented in C++
• Optimized spectral transformation for detection and classification of buried radioactive materials
• Proposed optimization methods to optimally combine the spectral channels by employing particle swarm optimization (PSO)
• Investigated the performance on suppressing spectral variants and enhancing the accuracy rate in later classification
• Modeled by Matlab and implemented in C++
• Anomaly detection of buried radioactive materials
• Implemented different anomaly detection algorithms to detect the anomalous spectrum of radioactive materials
• Applied 2D-interpolation to smooth the high frequency component around changing pixels for the probing area image
• Implemented in C++
• Estimation of mass and depth of buried radioactive material using neural networks
• Proposed PSO feature optimization for radioactive materials property prediction by using neural networks
• Studied and implemented backpropagation (BP), generalized regression neural networks and radial basis function (RBF)
• Implemented in Matlab

Teaching Assistant, 08/2007 - 12/2007

• ECE 3714 - Digital device
• Taught Lab, graded homework and exams, held office hours and tested digital device projects

Northwestern Polytechnical University (NPU), Xi\'an, Shaanxi, China
Research Assistant, 05/2006 - 07/2007

• Grey model in video coding
• Implemented a grey model to predict the pixel’s value and proved that this method was easier and more efficient than DPCM
• Implemented differential coding technique to reduce the relativity of the coded data

Member of NPU Mathematical Modeling Team, 07/2005 - 07/2007

• International Mathematical Contest in Modeling
• Meritorious Winner, The Mathematical Contest in Modeling (MCM), awarded by Consortium for Mathematics and Its Applications (COMAP) and National Security Agency(NSA) in 2007
• Attended the contest with thesis “Taming the partisan gerrymander to fair play"
• National Mathematical Modeling Competition
• First Prize, National Undergraduates Mathematical Contest in Modeling, awarded by China Society for Industrial and Applied Mathematics (CSIAM) in 2005 (2% awarded)
• Attended the competition with thesis “Applied water model to evaluate water quality”

SKILLS

• Platforms: Linux, Windows, Mac OS
• Languages: C/C++, Matlab, Python, Java, Visual Basic, HTML, XML, SAS, Shell scripting
• Databases: SQL, MySQL
• Tools: Visual Studio, gcc, g++, NetBeans, Qt, STL (Standard Template Library), MPI (Message Passing Interface), GPU (Graphics Processing Unit), CUDA, OpenCL, Intel TBB, Microsoft Office (Word, Excel, PowerPoint), Open Office

AWARDS AND ACTIVITIES

• Institute of Electrical and Electronics Engineers (IEEE), student chapter, 2009 - present
• Graduate Assistantship, Mississippi State University, 2007-present
• Scholarship, Northwestern Polytechnical University, three consecutive years, 2004-2006