Home Project Weekly My Current Fellow Final Report
First week...
When
I arrived in the laboratory, on June 7th, I first met my colleagues
at the laboratory and my mentor Holly Yanco too. She
explained the current project she was working with, and also talked about the
various projects she is involved in. She gave me the opportunity to choose the
project that I was most interested in. I
worked with algorithm in the first week and was able to learn about the
importance of robots in society and in the world including, for example, how
they are used in disasters, rescues or other emergency cases. I saw videos and
pamphlets about different models of robots and different uses. We had a meeting
and talked about the future work, goals and possible problems that I could
face. A graduate student talked to me and explained about the commands and task
development in each project.
Second week...
This
week was very important because we could choose the most interesting project
for us. We had a second meeting and talked about our points of view and
suggested about the projects. We had two visits this week from
Third week...
06/ 21/2004 - 06/25/2004
A little background about my project, some goals and point of view, my project is about the creation or improvement of an algorithm for a tracking system that can make a robot cross a street. Maybe it is more logical if the robot can move up to the nearest street light. Matters become more complicated when there is not a street light in the area, assuming that the robot is built with sensors, security systems. cameras, etc. When the tracking system works perfectly, the robot is able to track vehicles from a safe position. It may make tracking on speed, motion, and the distance of each vehicle approaching the robot. Once there is traffic in the street, it is more complicated for the robot to cross the street because its sensors detect small distances and slower speed. The robot could have a subroutine that could be to check around for a human in a position to cross the street, and it could imitate and follow the movement and speed of the human, moving in the same direction. Of course, it is possible that the person in a vehicle might not see the robot, and may not be aware that he is waiting to cross the street. One possible solution is to provide the robot with some indication that the driver of an approaching vehicle can see it and will give the robot an opportunity to cross.
Fourth week...
06/28/2004 - 07/02/2004
This
week I worked with the Phission code. I first needed
to know its principal function within my project. Phission
is a kit used to create and process subsystems in software. It is written in C
and C++ language. It is portable, efficient and fast. Basically Phission has three active components that are:
A Capture class: For
capturing images.
A Pipeline class: For
processing images.
And a Display class:
For exhibiting images.
The image is caught for a camera in the robot and the individual’s characteristics are based on color and are processed through a code created in Phission. Images are filtered through this code to learn the value of each color in each pixel. The RGB composition is used for this. One can also prove how this code works depending on the different values and colors applied to the code and filtered through the camera.
Another
exciting thing that happened this week was that I saw a project in the lab
about interface. A graduate student has been working on this project, which I
found extremely interesting.
Fifth week…
07/06/2004 - 07/09/2004
This
weekend, we celebrated the Independence Day. This was the 4th of
July weekend. I did not work on Monday and traveled to
I spent a lot of time reading more about the RGB Composition related to images and values and bits defined in a pixel.
Sixth week…
07/12/2004 - 07/16/2004
I
worked in the lab with the Phission code. The
graduate student who works with me gave me a part of this code and my mentor
explained the principal functions of the program. In RGB the composition is based on three different images of
the same object. The difference is that the light is filtered through a device
before the image can be detected. There are three filters: one for each color
(three basic colors- Red, Green and Blue). The images of the same object in
different filters are not same. The RGB assigns each pixel of the image a value
between 0 and 255 that represents the number of tones of the three basic
colors. RGB is not a good method for the
detection of a tracking system. It is good for the human eye, but not very good
for a camera. One problem with detect images in RGB is that when there is too
much light it's absorbed for the object and this causes distortion in color and
changes the value of the pixels of the image detected, too cause shadow
and all this can cause complications in the tracking system.
Seventh week…
07/19/2004 – 07/23/2004
I worked out with Mike different problems related to
the street crossing system. I learned how to work this system in real time. We
are working to improve the system and make it more efficient because there are
some problems that that need to be solved. We have a camera located in a point
in the street close to the lab, in the same place where the robot will be. It
is connected to a system in the lab and we will be able to see how the street
crossing system works. I learned how to track the vehicles with the camera. The
idea is to lock up in boxes all the objects in motion such as cars, bike,
pedestrians that are approaching or moving away from the camera in a direct
line. If the object in motion crosses right or left the box or frame disappears.
Another problem is the shadow underneath a vehicle because it causes oversize
boxes and this (box) becomes a limitation for the track.
Eighth week…
07/26/2004 – 07/30/2004
This week was awesome. We traveled to the AAAI
Conference in San Jose California. It was a great learning experience. Some
people from the lab worked as members in a team robot in the competition of
human rescues. This was a simulation of
a place destroyed by a disaster, and various team programmed robots had to find
survivors. Some robots worked great, and some were very slow. We also attended
a workshop on artificial intelligence. We finished the Conference with a party
given for Google. The conference was a lot of fun.
During our free time we walked around the city.
Ninth week…
08/02/2004 – 08/06/2004
When we returned from California, we had a meeting in
the lab with my mentor Holly, and we talked about our experience in California
and about our task in this week. I worked with Mike with a better image for
street crossing. I have to do a
presentation the next week about my work in the lab and Mike is helping me to
prepare the material.
Tenth week…
08/09/2004 – 08/13/2004
I can’t
believe this is the last week. I will be
all week preparing for the presentation. I am excited because I am going back
to my house and I want see my family. I feel a little sad at the same time
because I am going to miss to people in the lab and my mentor too, but I feel
everything went very well and it’s really been a great experience.