skip to main content

Course Websites

CS 445 - Computational Photography

Spring 2021

Official Description

Computer vision techniques to enhance, manipulate, and create media from photo collections, such as panoramic stitching, face morphing, texture synthesis, blending, and 3D reconstruction. Course Information: 3 undergraduate hours. 3 or 4 graduate hours. Prerequisite: CS 225, MATH 225, and MATH 231.

Related Faculty

Course Director

Text(s)

Varies by semester

Learning Goals

Be able to think about and analyze images in the spatial domain (1)

Be able to think about and analyze images in the frequency domain (1)

Be able to construct simple image filters and apply them to images (1)

Be able to convert to different color spaces and describe the advantages/disadvantages of each (1)

Be able to describe/compute how surface orientation, materials, light position and brightness, and camera position affect recorded intensity (1)

Be able to describe image intensities with histograms and adjust image intensity distributions to improve contrast (1)

Be able to implement template matching (1) (2)

Be able to implement texture synthesis as an application of template matching (1) (2)

Be able to construct graphs for analyzing images, for example in boundary finding or image segmentation (1)

Be able to transform, warp, and blend images to register or combine them (1)

Be able to compute transformations between pixel coordinates and 3d rays given camera parameters (1)

Be able to measure objects in images based on camera parameters or vanishing points and reference objects (1)

Be able to construct high dynamic range environment maps from photographs of a mirrored ball(1) (2)

Be able to automatically align images using interest points (1) (2)

Be able to apply computer vision and computational photography techniques to a topic of choice (1) (5)

Be able to present results informally via a web page (3)

Be able to present results formally via a poster and paper (3)

Topic List

Basics of Working with Images: filtering, frequency domain, histograms, color transformations

The Digital Canvas: Coloring, Blending, Cutting, Synthesizing, and Warping Images

Modeling the Physical World: Camera models, 2D / 3D projections, light capture and modeling

Correspondence and Recognition: feature matching, alignment, recognition

Advanced topics: multiple state-of-the-art methods that demonstrate the application of concepts learned in class

TitleSectionCRNTypeHoursTimesDaysLocationInstructor
Computational PhotographyC368994ONL3 -    Derek W Hoiem
Computational PhotographyC468997ONL4 -    Derek W Hoiem
Computational PhotographyDSO70855ONL4 -    Derek W Hoiem