Week of 24 January 2006
-
Administrative tasks: I turned in the Senior Expo
Intent to Participate paperwork to OTEFD. The current Expo date is 13 April
2006.
-
Laser Distance Meter: Over break, I looked into
the laser meter idea that Nick found (http://www.pages.drexel.edu/~twd25/webcam_laser_ranger.html)
using a laser pointer and a webcam and then calculating the distance by finding
the brightest pixel. So far all of the equipment works. I found a way to power
the laser pointer without a battery; next week it should be able to start
running off of robot power. Dr. Malinowski brought down a new webcam that will
auto-install itself on the laptop and screw onto the pan-tilt unit, and I was
able to capture a few images of the laser dot to start playing with. Next week
I will attempt to create a red filter for the camera to make the dot more
visible.
-
Pan-Tilt Unit: The pan-tilt unit (PTU) that will
be used to turn and tilt the laser meter assembly is now up and running off of
robot power. Dr. Malinowski helped configure HyperTerminal to work with the PTU
to send and receive data.
-
Plans for next week: Get the robot moving, and
continue with image processing.
|
|
Week of 31 January 2006
-
Image Processing and Capture: The camera has a red
filter now (red cellophane over the lens) to make the laser dot look brighter.
I am still trying to find a good way to get images from the camera into a
format that I can process in C++. The code from
http://www.pages.drexel.edu/~twd25/webcam_laser_ranger.html was written
for a Logitech Quickcam, which uses its own software and is not the webcam that
we're using for this project. Dr. Malinowski and I found a Microsoft
application that sounds like it does exactly what we want to do, but it only
runs on the most recent version of Visual Studio, which I don't have. There are
a couple more options that could be promising: Two previous senior projects
that used stereo cameras to track migrating birds developed software for
capturing frames from digital cameras. Two other ideas that Dr. Malinowski had
were DirectX from Microsoft and a library called TWAIN, both for image
processing. I am looking into the bird tracking project code first.
-
Plans for next week: Continue with image capture
and processing.
|
|
Week of 07 February 2006
-
Image Processing and Capture: The bird tracking
software used a frame-grabber card, which I will not be installing on the
laptop, since it'll just add complication and there's already enough equipment
coming out of the USB ports as it is. Nick suggested I look into the code from
the HOMERS project
of 2002 (Lovitt, Barngrover, and Knaub), since they essentially did the same
thing we're tyring to do, but with stereo vision and looking for a red ball. If
I can get their capture software working, then I can fill in my own
image-processing. They used the TWAIN library that Dr. Malinowski had mentioned
last week. Problem: The program requires a header file called twain.h, which I
had to download from twain.org instead of
their website, and the compiler won't recognize some of the typedefs it uses.
Next option: DirectX from Microsoft?
-
Plans for next week: Continue with image capture
and processing and try not to get further behind schedule.
|
|
Week of 14 February 2006
-
Image Processing and Capture: Switching to MATLAB Andy
Lovitt answered an email to tell me that the TWAIN interface was essentially
depreciated at the time that he was using it, and in any case, our camera would
have to have TWAIN drivers installed, which it doesn't. I talked to Dr. H., who
mentioned that Rob Scherbinske used the MATLAB Simulink Image Acquisition
Toolbox and Image Processing Toolbox on
his group's project last year. Since I'm running low on options, and
since the whole image processing part of this project is taking very much
longer than I had originally planned, I'm willing to go with this approach. The
disadvantage is that MATLAB is slow to run. The advantage is that I actually
prefer MATLAB to C++ for the image processing. Maybe I'll be able to get back
on schedule! I created the Simulink model for the image capture process, and
removed the red cellophane from the camera. It is no longer needed, and is
probably distorting the image more than it ever helped. The image is captured
in *.jpg format, which means that there are three color plans (red, green,
blue) when it gets through Simulink.
-
Next issue: How to communicate between MATLAB, the
PTU, and the robot?
|
|
Week of 28 February 2006
-
Administrative Issues: Missed last week due to
illness. This week Thursday morning I have a project progress report
presentation. I am now officially behind schedule.
-
Image Processing: I'm trying to work out an
algorithm to identify the laser dot. This is a problem because there are
several pixels with the brightest value in the image. Windowing may help, but
the laser pointer and camera will have to be permanently placed for a window to
be established.
-
Plans for next week: Continue with Image
Processing, look into mapping software (reference
http://www.jonh.net/%7Ejonh/robots/mapping/submitted-paper.html)
|
|
Week of 7 March 2006
-
Image Processing: On hold until the system is
mounted on the robot. Hopefully this will happen over spring break.
-
Mapping: Due to the inaccuracy of the digital
compass in the presence of EMI (building, robot), it won't be used in the
project. The ultrasonic sensors may be used though to detect objects at very
close range in an emergency stopping routine. I developed a set of dummy data
for creating a test map in MATLAB. I'll be using a probabilistic approach (no
obstacle = 0, definite obstacle = 1, maybes somewhere in between), and trying
to grow the map as new data is taken.
|
|
Week of 14 March 2006
-
Spring Break: Nick Schmidt and Mr. Mattus mounted
the PTU, laser pointer, webcam, and laptop frame on the robot while I was gone.
It looks even cuter than it did before, not to mention having all of the parts
connected and powered by the robot really helps to keep the project
progressing.
|
|
Week of 23 March 2006
-
Laser Meter: I took measurements to calibrate the
distance meter, following the
Drexel web page, and came up with an initial equation. Later, when I
tried to take long-distance measurements, I discovered that the laser pointer
was tilting upwards. After Nick fixed that problem, I added the new data to the
old set, and came up with a slightly different equation, which appears to work
very well.
|
|
Week of 30 March 2006
-
Administrative Issues: This week I had a practice
presentation with the speech coach, and there are several other presentations
and deliverables due shortly. The next big event is the Student Expo, coming up
in a couple of weeks.
-
Servers: Dr. Malinowski wrote a server to control
the PTU that I tested this week. Shom helped me set it up, and it runs through
telnet. The robot server has issues with the robot software. Saphira is no
longer supported, and doesn't work with newer operating systems like XP.
-
Mapping Software: I spent some time working on the
mapping routine, setting up the initial map array and colormap, and converting
from polar to cartesian coordinates.
|
|
Week of 6 April 2006
-
Administrative Issues: Expo is coming up, and I
need to get posters printed off ASAP.
-
Robot Software: Since Saphira is no longer
supported, ActivMedia sent Dr. Malinowski a password for the newer version,
called Aria, which he is trying to get working wiith a server.
-
Mapping Software: Meanwhile I am still trying to
keep the mapping software progressing. This week: growing the map to accomodate
new obstacles.
|
|
Week of 11 April 2006
-
Expo Week: MapBot and I went to Expo on Friday.
Check out the photos! It went really well, and we took a
few mappings and got to talk to a lot of people who came through.
-
Servers and Software: The PTU server can now be
controlled through MATLAB using an IP Toolbox created by Peter Rydesater. The
program maps completely once, but has a few minor issues with things like a
mirror image effect and some data points that could be classified as wild
anomalies.
|
|
Week of 20 April 2006
-
Tuesday: My practice presentation with the speech
coach was cancelled, so I worked on fixing the mapping software instead. It
seems to work pretty well now.
-
Wednesday: MapBot and I took a field trip to the
Institute for Learning in Retirement at the Student Center. He was a hit, and
he got to wear his 2nd Place ribbon from the Expo. :-)
-
Thursday: Working on fine-tuning the mapping
software to allow for multiple mappings, i.e., the robot takes a map twice in
the same place, and plots them on the same map.
|
|
Week of 25 April 2006
-
Administrative Issues: Final Presentation
Thursday, Poster Presentation for alumni on Friday.
-
Robot Server: Dr. Malinowski came down to the lab
and set the server up the right way (so that it didn't give the 88 linking
errors I was getting...) and the robot moves around! We had a great time almost
running into the professors on Friday during the poster presentation. :-)
|
|
May 8 2006
-
Demo Day: I got the robot running from a remote
PC, took a mapping, and played around for a few hours before the Demo. Then,
when I switched from manuevering to mapping mode, the computer crashed. When it
came back up, it was reading the serial port as a mouse - not good for trying
to run anything else. So MapBot's performance at the Expo will be accepted as a
demo instead.
-
Wrap Up: The final reports and presentations will
be posted on the website. I've commented all of my code, and I hope to post
that on the website too, just in case anyone wants to build off of the project
next year.
|
|