Computational Computer Vision

(3-0-0-3)

CMPE Degree: This course is Selected Elective for the CMPE degree.

EE Degree: This course is Selected Elective for the EE degree.

Lab Hours: 0 supervised lab hours and 0 unsupervised lab hours.

Technical Interest Groups / Course Categories: Threads / ECE Electives

Course Coordinator: Patricio Antonio Vela

Prerequisites: ECE 3084 [min D] or ECE 3550 [min D]

Catalog Description

Computational and theoretical aspects of computer vision. Application areas include robotics, autonomous vehicles, tracking, and image-guided surgery. Includes major project.

Textbook(s)

Robot Vision

Course Outcomes

Derive and describe image formation with regards to world and camera relative geometry;

Describe the process of camera calibration and its related optimization formulations;

Apply stereo triangulation and epipolar constraints for solving vision related problems;

Describe the relationship between the heat equation, diffusion, and Gaussian smoothing;

Apply and implement differential and convolutional operators as discrete stencil operations;

Explain the role of optimization in solving vision-based problems or estimating visual properties; 

Describe the purpose of adding prior constraints or regularization terms to computer vision derived optimization problems; 

Describe fundamental approaches to segmentation and clustering; 

Explain the equations underlying optical flow and their derivation from cost functionals;

Strategic Performance Indicators (SPIs)

N/A

Topic List

  1. Introduction
  2. Classic methods of linear filtering
  3. Sketch of wavelet techniques
  4. Edge detection
  5. Segmentation
  6. Optical flow and stereo disparity
  7. Shape recognition
  8. Color and texture processing
  9. Motion planning and tracking
  10. Applications: Robotics, Image-guided surgery, Controlled active vision