Yuxiao Zhou
EXPERIENCE
Fugro Roadware
Software Developer
Design and develop new Computer Vision algorithms for a multi-sensor mobile road analyzing platform (ARAN). Work in an agile environment and deliver new features for Vision, an all-in-one data processing and analysis software product. Responsible for most aspects of software development, including research, design, implementation, documentation and tire 2 support.
Computer Vision
- Designed and implemented computer vision algorithms to classify various pavement features from range images (in C++, C#, OpenCV).
- Designed and implemented image processing algorithms to detect various pavement features from 2d images with heavy noise, which achieves True Positive Rate of 90% (in C++, C#, OpenCV).
- Optimized existing panorama stitching work flow by applying a local nonlinear color and luminance compensation algorithm to create high-resolution and high-quality panorama where the source images have very different colors and luminance. (in C++, C#, OpenCV, OpenMP)
- Accelerated and refactored the implementation of the models for pavement surface 3d reconstruction. (in C++, C#, OpenCV, OpenMP)
- Designed and implemented novel methods to evaluate the performance of developed computer vision algorithm. (C++, OpenCV, Java, ImageJ)
- Created tools for acquired image quality analysis. (C++, OpenCV).
Data Analysis
- Developed tools for data extraction, manipulation and integration from various platforms using SQL and Java.
- Utilized genetic algorithm for data prediction model calibration and validation (in Java and MOEA framework).
- Built performance predictive models (Polynomial Regression, Logistic Regression) using Python,Numpy and Scipy.
Education
- M.S. in Computer Science, 2015, Jackson State University, Jackson, MS. GPA=4.0
- M.E. in Architecture, 2011, Shandong University, Jinan, China.
- B.E. in Architecture, 2006, Shandong University, Jinan, China.