User:Eshipley/my work

3D Scan Team Page

Project Goal

 * Use the Kinect to capture a 3D image and print it using the Maker Bot

My First Task

 * Reading all of the project's history which was started last semester, Fall 2011, by 501_cupcakes, and moving the most prevalent

information to the 3D Scan page on wikiversity so that it is more accessible will be my first task. I will research 3D scanning and explore

the subsystems to the project.

Summary of actual work over first weekend
How is what you did different than what you planned?
 * First of all, I spontaneously created a team collaboration page in Google Docs that is accessible to my team members through Canvas,

the school's course management system. Then, I noted the accomplishments of the past teams that have worked on the 3D scan project.

Next, I started preparing a presentation about the issue of not having an easily accessible 3D scanning device here at Howard Community

College. Finally, I looked for inspiration on 3D scanning and tips on giving a good presentation.

Week1 Narrative

 * Last Thursday, I promised my team that I would review the project history and research to obtain a better understanding of the

challenges we face in completing the project. Before leaving, I started a Google Doc collaboration through Canvas, Howard Community

College's (HCC) course management system, so that my team could keep in touch quickly and easily. In the document I defined the project's

requirements in using particular subsystems. I wanted to make sure that the focus was on getting the Kinect to work with the Maker Bot

because the last team working on 3D scanning was unable to create a quick and easy way to scan, model, and print 3D objects. Similarly, I

wanted the processing stages to use a software that will be easy to use and extremely helpful to future students and engineers at HCC.
 * Upon starting my research, I found a full tutorial on how to turn the Kinect into a 3D scanner. After a while of researching, I had found

several images of objects that had been scanned and printed in 3D, instructional methods of working with model files, some of the recent

history on 3D scanning (there is not a lot, so HCC students should start writing it), and a couple fun/inspirational links to websites like

The Top Ten Best Kinect Hacks and 3D Printing Is The Cover Story Of USA Today. I shared all of this in the collaboration document to

give my team a better understanding of the project and inspire and motivate them to push the project forward.

My Second Task

 * This weekend I plan on using the tutorial here on Wikiversity to set up the MakerBot and print a 3D object file which I will download from

an internet archive. If I come across any obstacles in doing so, I plan on documenting them in my notebook, so that I can improve the tutorial.

Also, I sent a message to the software developer engineers at INUS Technology, Inc. the creators of RapidForm XOR about receiving a

demo copy of their state of the art unreleased 3D software and having a presentation from one of their engineers for HCC's engineering club,

and I have already received a personal e-mail from them that told me to expect a call from one of their engineers tomorrow. I plan on doing my

best to put these plans into action.

Summary of actual work over second weekend
How is what you did different than what you planned?
 * I left a message for the Rapidform developers, but haven't heard back. Over the weekend I re-installed the drivers for the MakerBot, and was

able to print out a cone shaped model with a few additional instructions from a student worker.

Week2 Narrative

 * In order to force the MakerBot to print at a lower temperature, I needed to change a line of code in the generated g-code. Notably, the new

necessary temperature setting is 180 degrees Celsius. I hypothesize that the reason for this is an error in the MakerBots calibration, since

new parts have been installed on it. And, the actual temperature of the machine and the one specified originally in the code is 220 degrees

Celsius. I learned that the MakerBot prints a reflection of the model that is seen in the ReplicatorG software by printing a key chain that has

the words; "Start Making." Thus, to print something and spell it from left to right, the words must be backwards in the software (reflected

across the x-axis.) I drew up some sketches of the subsystems that might attract attention if they were made into 3D visual props and/or

added to the project poster as drawings or pictures. At the end of the week, I installed AutoDesk Maya, watched the basic tutorials, and

became familiar with the controls and most of the interface buttons.

My Third task

 * For my third task I plan on learning how to clean up a point cloud in MeshLab, save it as an object file, and render it in Maya in

preparation for it to be printed. Also, I will organize materials for a project poster/backboard/exhibit.

Summary of actual work over third weekend

 * I did not make it to Office Depot or Kinko's to pick up presentation tools. I also was not able to work on a point cloud because I did not

have the tools to make one. Instead, I spent some time learning to use the MakerBot.

Week3 Narrative

 * I was unable to print with the MakerBot upon arriving, but I was able to troubleshoot the problem and found that the easiest solution

fixed the problem. Filament was not being extruded. Though, there are many reasons this might happen, the solution involved tightening

the thumbscrew pressing the material wire into the CNC motor. I outlined this and the steps I took when I had to re-install the drivers in

the MakerBot tutorials.

My Fourth task

 * Make almost done.jpg plan on taking pictures while pushing the project forward. This way I will have material to create a presentation without taking too

much time away from successfully getting a scanned object printed. I will work with the Kinect and processing software to get a point

cloud ready for Meshlab. Afterwards, I will edit and convert the file until it can be printed using the MakerBot.

Summary of actual work over fourth weekend
How is what you did different than what you planned?
 * Worked through every step of the project alone, including designing a turnatable.

Week4 Narrative

 * Throughout the week I was sidetracked with the MakerBot tutorials, my LCD side project, and organizing presentation materials for

the project. However, I came to school Friday afternoon and made a great deal of progress. Dr. Foerster rolled back one of the computers

to a Windows 7 32 bit operating system, so that I could push the project forward. Both software packages that I have found to create a

three hundred sixty degree three dimensional view are designed to run on the Mac OS or Win 7 32 bit OS. I began by downloading the

software packages. From this point on, I took screenshots of every step I took in preparation to build a tutorial here on Wikiversity. I

tested each of the execution files in the RGB demo. I started to realize that I was going to need some additional instruction to callibrate

the Kinect so that it would build a 360 degree mesh of the appropriate object. However, I found that I could still create a 360 degree view

of a symmetrical object such as a soda can as long as I centered it in the Kinect's focal point. I created a temporary turntable made out of

a cardboard box and a circular platform that had a rotating ball-bearing bracket attached to it's bottom side. Also, I put tape inside the

box, but that turned out to be an unnecessary step because the tape method or use of a checkerboard pattern on a paper is used when

calibrating the Kinect, which I did not do. When I scanned the soda with my new turntable I had placed the Kinect almost directly above the

composition, so my point cloud consisted of data points for solely the top of the can. Next, I quickly captured a point cloud of myself from

a frontal view. Then, I opened the mesh of myself which is saved as a polygon file (.ply) in MeshLab. I deleted the point cloud of the room

in the background and performed some sampling and reconstruction filtering to connect the dots and turn the mesh into an object that can be

saved into the file type used by the MakerBot for printing. However, some objects may require a few additional editing techniques before they

can be printed. For example the model of me was just a contour and did not have a base. After watching several more youtube videos, I have

come to the conclusion that the better software for creating a 360 degree view of an object is ReconstructMe. I plan on making a tutorial

related to the subject matter seen in this example video in addition to the 3D Scanning with RGBDemo tutorial (will create link here asap).

Complete Project Page
Follow the "project done" format.