yinxia39
Dec 1, 2013
Undergraduate / I was fascinated by how the Xbox Kinect worked; UMich: Curriculum support my interest [2]
Any feedback would be great. Be ruthless and rip apart my essay if necessary.
Prompt: Describe the unique qualities that attract you to the specific undergraduate College or School (including preferred admission and dual degree programs) to which you are applying at the University of Michigan. How would that curriculum support your interests?
As a junior in high school, I was fascinated by how the Xbox Kinect worked. It had a camera that had, as an added bonus, the capability to detect depth. The new games that are available on the Kinect are very user oriented, where movements of the user are the primary controller of the character in game. I wanted to know how the Kinect was capable of detecting, and interpreting the motion.
Although I knew I could find the answer online, I decided to try to do some motion detection myself. I hypothesized that in order to detect motion, I needed a way to describe an object based on what the computer can extract from an image. I began with something simple: a program that located the brightest location in an image, and circled it on screen. My approach to this was to sum up the pixel color value for each row and column. This way, the row and column that has the greatest sum would be the location of the brightest spot in the image. I first tested this method on a simple image, with a single white dot on a solid black background. When I got the program to circle the dot on the test image, I went ahead, and set my webcam feed as the image source. In contrast to my expectation, this method did not work well at all. When there is a noisy background, the "location" of the brightest spot in the image would jump around aimlessly.
"Why does this not work?" I asked myself. I realized that there are several cases where the pixel sum for each row or column can be similar. Just summing up the pixel values of each row/column is analogous to an average, where there are several factors to take into consideration. To fix the jumpiness, I kept track of the location of the previous "brightest point" and weighted color values of each pixel by their linear distance from the previous brightest point.
I continued this path by interning at the Language and Media Processing Lab at the University of Maryland over the summer. Prior to my summer research project, I thought that computer science was an independent field of study.. Throughout the course of this research experience, I came across several bumps that were not solely computer science issues. I often came across psychology, physics, mathematics, and even philosophy questions. I realized that computer science is an interdisciplinary field. It combines a wide variety of knowledge, and applies it algorithmically. Computer science is the interface between theory and application. Realizing this I became interested in a wide variety of sciences including physics and psychology.
A program that I have set my sights on at the University of Michigan is the "Computer Engineering" program. I believe this is perfect curriculum for learning how to apply scientific theory to reality.
Word Count; 484/500
Any feedback would be great. Be ruthless and rip apart my essay if necessary.
Prompt: Describe the unique qualities that attract you to the specific undergraduate College or School (including preferred admission and dual degree programs) to which you are applying at the University of Michigan. How would that curriculum support your interests?
As a junior in high school, I was fascinated by how the Xbox Kinect worked. It had a camera that had, as an added bonus, the capability to detect depth. The new games that are available on the Kinect are very user oriented, where movements of the user are the primary controller of the character in game. I wanted to know how the Kinect was capable of detecting, and interpreting the motion.
Although I knew I could find the answer online, I decided to try to do some motion detection myself. I hypothesized that in order to detect motion, I needed a way to describe an object based on what the computer can extract from an image. I began with something simple: a program that located the brightest location in an image, and circled it on screen. My approach to this was to sum up the pixel color value for each row and column. This way, the row and column that has the greatest sum would be the location of the brightest spot in the image. I first tested this method on a simple image, with a single white dot on a solid black background. When I got the program to circle the dot on the test image, I went ahead, and set my webcam feed as the image source. In contrast to my expectation, this method did not work well at all. When there is a noisy background, the "location" of the brightest spot in the image would jump around aimlessly.
"Why does this not work?" I asked myself. I realized that there are several cases where the pixel sum for each row or column can be similar. Just summing up the pixel values of each row/column is analogous to an average, where there are several factors to take into consideration. To fix the jumpiness, I kept track of the location of the previous "brightest point" and weighted color values of each pixel by their linear distance from the previous brightest point.
I continued this path by interning at the Language and Media Processing Lab at the University of Maryland over the summer. Prior to my summer research project, I thought that computer science was an independent field of study.. Throughout the course of this research experience, I came across several bumps that were not solely computer science issues. I often came across psychology, physics, mathematics, and even philosophy questions. I realized that computer science is an interdisciplinary field. It combines a wide variety of knowledge, and applies it algorithmically. Computer science is the interface between theory and application. Realizing this I became interested in a wide variety of sciences including physics and psychology.
A program that I have set my sights on at the University of Michigan is the "Computer Engineering" program. I believe this is perfect curriculum for learning how to apply scientific theory to reality.
Word Count; 484/500