CS6420 -- Presentation Schedule

Roger Alms: May 04, 2015
Ben Davis: JPEG Encryption for Image Rescaling in the Encrypted Domain; April 22, 2015; This project implements a method to encrypt an image such that the original image can be retrieved even if the image is resized. In a number of instances, the original [large-sized] images are encrypted to thumbnail sizes to save space. We will present a method to retrieve the thumbnail image and decrypt it back to a downsampled version of the original image.
Robert Meadows: Least Significant Bit Steganography in Digital Images; April 27, 2015; Security has never been more important than it is now. As a result, we cannot rely on just pure encryption; we must also consider other forms of security such as information hiding. Encryption can draw attention while information hidden inside seemingly normal traffic will be ignored. This project focuses on implementing code for steganography using a technique called least significant bitg or 1-LSB. The goal is to produce a cover image that is unmistakable from the original source image while hiding the contents of another image. The project involves the implementation of a decoder to retrieve the secret image.
Drew Ryder: Analyzing Plant Root System Architecture; April 29, 2015; Roots are essential to a plant's function. Yet, relatively little is known about plant root system architecture. Several methods of root imaging have been developed to gather data on the root systems. However, there is a need to extract relevant phenotypic information from these images. My research focuses on implementing a program that analyzes plant root images to determine quantitative traits of the root system architecture.
Nathaniel Schremp: May 06, 2015
Tjaart van der Walt: Image Composition for Photographic Lighting; April 20, 2015; This presentation discusses techniques to support a new workflow that is emerging for the photographic lighting of static scenes. Imagine that a photographer wants to take a studio photo of a bowl of fruit. Traditionally, he will make an effort to set up the perfect lighting conditions. Then, he takes a perfectly lit photo. In recent years, some photographers have been using an alternative workflow where they take multiple photos from a fixed vantage point with a moving light source. They then use image manipulation software to merge these images into one well-lit image. This process requires a lot of time and advanced image manipulation skills. My presentation explains proposed techniques to simplify this process.
Joel Waymack: Facial Recognition and Identification with Eigenfaces; April 15, 2015; The recognition and identification of human faces is a field of increasing importance within artificial intelligence and image processing. This project investigates the use of eigenfaces image processing algorithm to correctly identify faces within a given image using principal component analysis and a large database of faces. The project also analyzes various approaches to try and increase facial identification success rates.