It can be accessed with brief instructions HERE

Also as part of the project i learned how to use GIT so the source of the project can be found on Git-hub HERE for those interested, so if you have any ideas or find any bugs that would be the place to leave them.

I hope somebody finds this useful, and if not i think i may have some uses for it in future projects (such as super multiplayer multilayer networked pong)

---

Add to

However i started to run into problems when i needed to be able to split sections of the page into columns. More extensive googleing didnt turn up anything, and the best i found that was in rtf documents the command \colsN where n is how many columns needs to be passed in with the section settings.

Searching the source code for cols didnt turn up anything so,  after much frustration caused by having a copy of PyRTF from a previous project in my site-packages directory, the functionality was easily added and my project is once again underway.

For those who need similar functionality in their projects and couldn't be bothered fiddling around with source code

Get PyRTF with my modifications here

---

Add to

To open it you press the button on either side

I created this tool out of a piece of coat hanger to press in the ends of the spring pins to remove the clasp from the watch

The clasp mechanism after being removed from the band and its backing by removing spring pins with the above tool, The next stage was to unbend the four pins keeping the cover on, i forgot to get a photo after i had taken that off. (And yes the grime on it is quite disgusting)

As soon as I took the panel off a few little springs bounced out, and as i suspected they were broken and tangled. The photo is off one of the largest pieces next to a 5c coin for reference

This is the socket that the spring was supposed to rest in, unfortunately I didn't get an overview of the mechanism, but there was another similar socket on the other side

And here was the cheap and simple fix, having no springs anywhere near that size. I cut a tiny fragment off an elastic band.

The only remaining thing to do was to reassemble the clasp and re-attach it to the band and it worked, it now no longer falls off by itself. I'm not sure how long the rubber will hold up but until it fails im happy.

The closeup images where taken using the setup HERE

---

Add to

A section of my desk pad where I have scribbled to get pens (standard ballpoint) working

Closer up of the pad, you can even see the fibers in the paper

Close up of some 2mm braided cord

As a last test out of interest I took a photo and LCD screen, the lines are so bright you cant even make out what was originally on the screen

Last of all here is the setup I used, all it is is the lens blu-tacked to the front of the camera, not shown is a piece of post it note over the flash to give a less glared image

Edit: I have also used this setup to take photos of some of the fine parts in my watch when I did repairs, this can be found HERE

---

Add to

I am getting close to my goal of taking a gigapixel image, this time it was taken on a tripod with a more powerful lense. It was made from joining around 300 photos and I managed to avoid missing any bits out this time.

For this one I managed to use a latter version of photoshop at school to crop it a bit and remove most of the jagged edges.

Check it out HERE

Pointless to have a thumbnail of the whole thing, so here is a thumbnail at 1/4 zoom at the steel works

Slouching around taking a few hundred photos on the edge of a cliff

I hope you enjoy it, and find some interesting spots, I plan on making my own version of the zoomify viewer where you can mark locations, so it makes it a big game of i-spy.

---

Add to

Click for a larger version

A view from the large belay shelf on the bard

A view of the camp site, just after ali barbers cave

A different than normal perspective, looking towards Arapilies from Miter rock

Bard buttress and Tiger wall

---

Add to

The project consisted of:

• Building a turntable to be controlled by the computer that can rotate the object accurately
• Convert a laser pointer to emit a line onto the object rather than a dot
• Have some way to take pictures of the position of the laser line on the object
• Write programs to control the turntable, find and process the laser line in the captured images and display the processed information as a 3d model

Here is a picture of the general setup I ended up using (I didn't get a good photo of the new stepper and gearing I got to rotate the object)

In this image you can see the turntable (black circle), the laser pointer being refracted into a line using the glass rod in the top right, the circuit board used to control the motor, and the board used as a quick way to trigger the camera.

Here are some images from the different stages of the process:

The object to be scanned

The object sits on the turntable and is rotated accurately using a stepper motor. The first stepper I used allowed 98 steps per revolution (the scan you can see below) while the second stepper motor I used had extra gearing and allowed 290 steps per revolution.

The hardest stage in the entire project is the extraction of an exact line from the image. The stages of this are:

1. Load the raw image
2. Extract the Red channel out of the RGB image
3. Find the entire line in this image as a 2 bit image
4. Decide upon an exact line through the previous blob ignoring noise

Coordinates are calculated for each point using how high in the image each point on the line is, what rotation the table was at in the image and some basic trigonometry to create thousands of data points that are then linked and covered with faces by a blender script to create a model.

The Raw Model

Then I have simply manually tidied up the model removing where the table appears in the scan and smoothing it.

Trimmed and smoothed

Then I have applied a stock wood texture to the model to produce the final result. I think that it has turned out rather well for a first go at creating a 3D scanner especially as it only took a few weeks and cost me nothing.

The finished product

I am happy to answer any questions you might have and last of all if you are still curious here are some of the sheets I handed in with my folio to show how the points were converted after extraction.

---

Add to

One of the more interesting things that I had a go at was creating a really, really high res image. The one below consists of over 200 photos and when joined results in a photo around 400 Megapixels.

The main problem after you have waited for your photos to be stitched is how to edit them. Standard image viewers refuse to open them, photoshop crashes. Eventually I found a viewer that splits the images up into thousands of tiles that makes it easy and fast to view. I am still looking for a way to edit the finished image though.

I was very annoyed when I got home to find that when taking the pictures I had missed a narrow strip across the image. I plan on having a another go of taking more photos like this closer to home and I have ideas for a simple robot to automate the taking of them.

You can explore it HERE

How many climbers can you find?

---

Add to

(Click for a larger image)

---

Add to

http://code.google.com/p/google-appengine-wx-launcher/

It is GUI that has been created in python useing the wxWidgets framework and looks and acts very similar to its mac counterpart

For those who don't want to have to bother dealing with SNV or want an exe version i have compiled the SNV version (as of the 26-2-10)  and it is available here

I do not claim to have created the program at all, only compiled it for easy use using GUI2Exe, i hope it comes in handy

---

Add to

Since i had so much trouble with it i thought it would be worthwhile to post the solution to internet explorer corrupting some files that are loaded through PHP e.g.

header('Content-Description: File Transfer');
header('Content-Type: application/octet-stream');
header('Content-Disposition: attachment; filename='.basename($path));
header('Content-Transfer-Encoding: binary');
header('Expires: 0');
header('Cache-Control: must-revalidate, post-check=0, pre-check=0');
header('Pragma: public');
header('Content-Length: ' . filesize($path));
ob_clean();
flush();
readfile($path);

The simplest way i found to correct the problem was to comment (or remove) out the line declaring the length of the file

//header('Content-Length: ' . filesize($path));

The downside to this method is that for downloading large files, users wont know the total size of the file being downloaded, luckily for me the files been downloaded were small enough for this not to cause problems.

I hope that by writing this i manage to save somebody some time and frustration, leave a comment if you have any other fixes for internet explorers many problems.

---

Add to

http://www.youtube.com/watch?v=GmtTHkNd_Q0

It shows off more of the programs features for improveing accuracy and its ablility to repetitively enter rounds, answer a certain number of questions and then strike out and repeat.

---

Add to

Click for larger image

And this panorama after we built a giant sand pyramid, we calculated it and it turned out to be 10 tons 7 tons (correction) of sand (4.05 cubic meters)

Not very happy with this one picure

---

Add to

Click for a larger view

---

Add to