- <div style="background-image:url(/live/image/gid/84/width/1600/height/300/crop/1/29613_carolyn_tuttlecropped2.rev.1451944800.jpg)"/>
- <div style="background-image:url(/live/image/gid/84/width/1600/height/300/crop/1/29612_judy_dozier_2cropped.rev.1451943820.jpg)"/>
- <div style="background-image:url(/live/image/gid/84/width/1600/height/300/crop/1/29614_ben_zeller_2.rev.1451944512.jpg)"/>
- <div style="background-image:url(/live/image/gid/84/width/1600/height/300/crop/1/29615_rob_baadecropped3.rev.1451944211.jpg)"/>
- <div style="background-image:url(/live/image/gid/84/width/1600/height/300/crop/1/5986_screen_shot_2016-11-30_at_3.27.37_pm.rev.1480543045.png)"/>
- <div style="background-image:url(/live/image/gid/84/width/1600/height/300/crop/1/29616_ann_robertscropped2.rev.1451944021.jpg)"/>
- <div style="background-image:url(/live/image/gid/84/width/1600/height/300/crop/1/29610_alex_shingeltoncropped2.rev.1451943950.jpg)"/>
Mikita Zhylinski ’20
In today’s rapidly advancing world, tampering with digital images is becoming more of a common practice. Mikita Zhylinski ’20 is working on building a program to detect these tampered images.
With the help of Assistant Professor of Computer Science Sugata Banerji, Zhylinski is refining his computer science skills and looking at tampered images through his program to determine what has been doctored from one image to the next.
Q: Tell us about your research project.
A: “It is a program that defines the same parts on an image. One of the images is tampered and one is not. The program will be able to determine which part of the image is tampered with and copied. It uses MATLAB, which processes data as matrices and as mathematical functions, as the coding base.”
Q: What does it mean when a photo is tampered?
A: “Tampered means changed. For my program, it means that part of an image was copied and pasted over another part of the same image. On a tampered image, the program will show us where there are copied parts. The program takes each part and compares it with all of the other parts to see if there are any that are the same. Our idea was that if you take each pixel, you also take the surrounding pixels and you store the surroundings of this pixel in a vector, which is a line of numbers. So each pixel will be converted into an eight-size vector. This image has 424,081 pixels so each photo is a three-dimensional matrix. It is not a single matrix, because it is actually a cube. The pixel that interests us is in the center and the 75 pixels around it.”
Q: How will it benefit you to work with a professor this early in your academic career?
A: “I hope it will benefit my resume and career path. Working with the professor gives me the opportunity to learn about real life problems during college. It helps to be prepared for the real life.”
–By Tracy Koenn and Sophie Mucciaccio