Eduardo R. Corral-Soto
Ph.D. Candidate, M.Sc., P.Eng.


Eduardo Corral

Research Interests: Computer Vision Algorithms (Motion Analysis and 3D Reconstruction) and Image/Video Processing Algorithms.

Current Research:

  • I have been working on two projects at York University. In the first project which is called 'The GEOIDE PIV-17 Project' we have developed a near-real time 3D surveillance/monitoring system in which we augmented a 3D model of York University with animated avatars (e.g. pedestrians and cars) representing the dynamic information from the scene captured by live video cameras and sensors. My specific participation in this project was on the 2D-to-3D mapping via projection camera matrices and the tracking and back-projection of moving objects. Here's a link to a related Video.The second project is about developing automatic methods for the extraction of curves of interest from highway imagery using Bayesian inference, and algorithms for image rectification and camera pose estimation for arbitrary curvature highways without the use of vanishing points.

  • Education:

  • Ph.D. (In progress) Department of Computer Science and Engineering
  • York University, Canada.
  • M.Sc. Electrical and Computer Engineering.
  • University of Manitoba, Canada (Thesis: "Real-time Imaging System for a Ground Penetrating Radar").
  • Specialization in Embedded Computer Software Engineering.
  • University of California, U.S.A.
  • Bachelor in Electronics Engineering.
  • Instituto Tecnologico de la Laguna, Mexico.


    Teaching:
  • Winter 2012-CSC320H1S Introduction to Visual Computing (University of Toronto):
  • CSC320H1S Course website

    Publications:
  • Corral-Soto, E.R., Tal, R., Wang, L., Persad, R., Chao, L. Solomon, C., Hou, Y., Sohn G., Elder, J.H. (2012) 3DTown: The Automatic Urban Awareness Project, Proc. Canadian Conference on Computer and Robot Vision.
  • IEEE Canadian Conference on Electrical and Computer Engineering. Winnipeg,Manitoba, Canada. May, 2002. Paper: Effects of Using Superresolution Techniques in ISAR imagery.
  • AMEREM 2002 International Conference. U.S. Naval Academy, Annapolis, Maryland, U.S.A. June, 2002. Paper: High Resolution Imaging Using a SFCW GPR.
  • (Poster) E. Corral. CVR 2009 International Conference on Vision in 3D Environments, Toronto, Canada. Poster: Extending The Range Capabilities of a KLT-Based Video Stabilizer via Phase Correlation.
  • (Poster) E. Corral and J. Elder. Probabilistic Detection and Grouping of Highway Lane Marks. Presented at the GEOIDE 2010 Conference in Calgary (June 2010).

  • Patents:
  • Global Motion-Adaptive System with Motion Values Correction with Respect to Luminance Level. Genesis-Microchip Patent No. US 7,471,336,B2.
  • High-Quality Single-Frame Superresolution Training and Reconstruction Engine (Patent application).
  • Perceived Depth Adjustment Engine for 3D-TV (Patent application).
  • Adaptive Sub-pixel Accuracy System for Motion & Disparities Estimation (Patent application).
  • DBIS: Depth-Based Image Scaling for 3D-TV (Patent application).

  • Relevant Courses:
  • Computer Vision (York)
  • Statistical Signal Processing (York)
  • Statistical Motion Analysis (York)
  • Computational Modeling of Visual Perception (York)
  • Foundations of Computer Vision (audited at UofT)
  • Digital Signal Processing (Manitoba)
  • Digital Image Processing (Manitoba)
  • Fourier Imaging (Manitoba)
  • Signals and Data Compression (Manitoba)
  • International Computer Vision Summer School 2010 (Sicily, Italy)
  • NCAP Summer School 2011 (UofT)

  • Contact
    Email:   ecorral@cse.yorku.ca