House App Competition and Vectorization!

House App Competition and Vectorization!

House App Competition

Results are in: PiPlotter has won the Congressional STEM Challenge within California's 7th District! It's been an awesome experience working with the app, image manipulation, and video presentations. While this is only the first year of the challenge, I hope many more students will get to experience the competition and app building process.

Vectorization

As you may be able to tell from the pictures that PiPlotter currently outputs, it is still thinking in terms of a raster. The image input finds the edges and traces a path where the pen should go, but that path is stored as lines from pixel to pixel, creating a jagged line effect.

The solution? Take these paths and turn them into lines/curves flowing through them. While I don't have any build for this ready at the moment, I have some rough ideas on how to implement this.
 - Here is a concept for the algorithm / process I will need to make for the pixel interpretation
 - Here is a great documentation of Potrace's bezier calculations that I will likely use as inspiration.

May 7: Version 2.0 and Competitions!

It sure has been a while Matt, what have you been up to?

In the past month, I have accomplished:

• Servo lifts pen off paper
• Raspberry Pi accepts connections and DrawData over local network
• A completely fresh Android App allows for pictures to be transformed into line drawings and vector data

 On April 11th, I took part in my high school's Engineering Challenge. It was a fun experience having my project judged along with around 100 other awesome teams and projects.

The PiPlotter received an amazing 173.75 / 175 and won first place in the overall competition. After the Engineering Challenge, I kept making forward progress with my code and finished another complete "final build" yesterday. Today I entered my finished product into the House Student App Competition.

April 13: 1st Working Build and Presentation

My pen plotter has been fully built! There are still lots of improvements to be made, but this is just the first build. My presentation was Friday, and many people were impressed by the creativity and hard work.

Mar 22: Source Code on GitHub

Source Code Posted on GitHub

I've been focusing mostly on the software aspect of the project for the past few weeks. If you wish to read up on my source code, it will be periodically posted on my GitHub page (https://github.com/MatthewMerrill/).

Mar 16: Wheels finished!

Well... okay halfway finished - Hobbytown only had 1 of the 2 bolt+locking gear pieces I needed, and the other location only had a longer version. Another will ship in Tuesday, but this wheel is ready to go!

Feb 22: Basic Rundown Video Uploaded!

Basic Rundown Video:

Here is a video explaining the basics of the physical aspect of the project. In this video, I...

• Disprove the Wadsworth Constant
• Explain the easel rig and how the components will be placed.
• Show off the stepper motors doing their thing
• Announce the arrival of those amazing sugary chocolate egg things

 Motor Component Explanations:

• Servo Explanation
• Stepper Explanation

Feb 13: Parts Arrived and Bézier Curve Rendered!

Electronic Parts:

The parts came in the mail Monday! Was very excited to test out the raspberry pi and got a "Hello World!" working in no time! No problems with the parts so far, everything has been looking up. I'll try to edit and post the unboxing / basic project explanation video sometime over the weekend.

Bézier Curve:

For the past couple days, I've been pondering how to tackle the problem of "tracing" a Bézier curve. After a few minutes on wikipedia, I realized that Béziers are a lot more straightforward than I had originally thought!

 The basics of a Bézier is to draw lines between the ends and the controls, then find a point at percentage "t" out. From these points you draw a line and continue out, etc. and repeat the process until you are left with one final point, which represents the position on the Bézier "t" percent of the way to the other end...

Wikipedia's "Constructing Bézier curves" has some great examples of how exactly this is done. The one that inspired me the most is this one.

After a couple of minutes and well under 50 lines of programming I cranked out the render seen above.