Research update March 13, 2018

This week I have no progress report due to preparing for final exams for the quarter. My research summary was accepted for the 2018 CSU Statewide Research Competition. I will have progress to report at the end of this week.


Research update March 1, 2018

This week I completed my abstract and research summary for submission for Grace Hopper Conference 2018. The title is “LLE Based Image Hashing of Video.”

This upcoming week I will continue working with the video datasets recently created to further conduct my experiment.

Research update Feb. 22, 2018

This week I read two papers. 1. Robust Image Hashing 2. Robust and Secure Image Hashing. I completed and submitted a research summary for consideration for the CSU Student Research Competition. I have collected more video data. I will also be getting video from another research group at my University. This upcoming week I will be finalizing a research abstract for submission to a conference. I will also be continuing with the video experiment on the slow motion video dataset.

Research update Feb. 14, 2018

This week I read two papers. 1. Hashing on Nonlinear Manifolds 2. Wallflower: Principles and Practice of Background Maintenance. I have began writing up a research summary to be submitted for an upcoming conference. This upcoming week I will begin testing an experiment on video data against two kinds of datasets, one with frames superimposed as part of the preprocessing and the other without superimposed preprocessing. The data is coming from slow motion video that I captured this week.

Research update Feb. 6, 2018

This week I read several papers. 1. Faster and Better: A Machine Learning Approach to Corner Detection 2. Effective Gaussian Mixture Learning for Video Background Subtraction 3. Robust image hashing via DCT and LLE. I also met with my mentor to discuss the project and to plan for the upcoming weeks.

1. In this paper they discuss the many approaches to detecting corners in images. Detecting corners is a common first step to analyzing image information for computer vision tasks. Roston, Porter, and Drummond provide an improvement to the FAST detectors by introducing the FAST-ER detector. They are able to use the FAST-ER detector on live PAL video.


Figure (Figure 1) and description from the paper: Twelve-point segment test corner detection in an image patch. the highlighted squares are the pixels used in the corner detection. The pixel at p is the center of a candidate corner. the arc is indicated by the dashed line passing through 12 contiguous pixels which are brighter than p by more than the threshold.”

2. In this paper they discuss subtracting background information from video data. In particular they investigate data from traffic surveillance videos and meeting videos. “To asses the impact of the learning algorithm in a real system, we applied mixture modeling to background subtraction using a statistical framework.” They were able to improve convergence rate and estimation accuracy based on current methods available in 2005.

3. In this paper they discuss image hashing which is used commonly in various applications that check to see if an image is a similar or visually close to another image such as in image authentication, image indexing, and image forensics. They propose a novel approach to image hashing via DCT (discrete cosine transform) and LLE (locally linear embedding). Their hashing performed better than GF-LVQ hashing, LLE-based hashing, QFT hashing, QSVD hashing, and DCT-NMF hashing algorithms in classification between robustness and discrimination.


Figure (Figure 2) and description from the paper: “Comparisons between the Euclidean distances and color vector angles of two color pairs.”


During the upcoming weeks, I will be preparing slow motion video datasets, conducting further literature reviews, and preparing research summaries for submission for two upcoming conferences.


Lee, D.-S. (2005). Effective Gaussian Mixture Learning for Video Background Subtraction. IEEE Transactions on Pattern Analysis and Machine Intelligence, 27(5), 827–832.

Rosten, E., Porter, R., & Drummond, T. (2008). Faster and Better: A MAchine Learning Approach to Corner Detection. IEEE Transactions on Pattern Analysis and Machine Intelligence, 32(1), 105–119.

Tang, Z., Lao, H., Zhang, X., & Liu, K. (2016). Robust image hashing via DCT and LLE. Computers & Security, 62, 133–148.