I started out by reading up on the Canny Edge Detect algorithm.  There are lots of sources of information, but the best outline seems to be on Wikipedia. http://en.wikipedia.org/wiki/Canny_edge_detector

For testing I just used a couple of the images that come with the windows install.

The first step is to blur the image so only the really prominent edges will be detected and hopefully any noise will be eliminated.  You can do this by convolving your image with a small filter matrix.  Wow, look at all the fancy terminology.  Essentially what this meant to me was that every pixel in your image needs to be blended with the pixels around it.  You can either go out one layer with a 3×3 matrix or 2 layers with a 5×5 matrix.  You can go bigger, but the the information I found seems to suggest that it is not necessary.  The matrix basically tells you how much weight each of the surrounding pixels should have on the new pixel color you are generating.  So the pixel you are looking at should have the biggest weight.  The first layer of pixels around it should have a slightly lower weight.  And if you go further out, each layer should have smaller weight then the one before it.  The picture below gives a visual representation of this idea.  There is a mathmatical theory for figuring out what those percentages should actually be, but you can probably find better explanations of it elsewhere.

Next, gray scaling the image helps in the calculation of what is actually an edge.  Instead of trying to calculate if the pixel at point X is a different RGB color than a pixel at point Y, you can simply use the gray scale intensity to make the comparison.  This drops you from a 3 value to 3 value comparison to a 1 value to 1 value.

Although to a human eye, edge detection is a very simple process, it is not so easy for a computer.  We have to perform a couple different passes to determine what is actually an edge and what is not.  I chose the Sobel operator which returns potential edges in the horizontal and vertical direction.

These images can be combined and some additional math performed to find the direction of each edge that was detected.

Using non-maximum suppression and tracing edges with thresholds can lead to some very good results.  However, in my case, I wasn’t as happy with the final products as I was hoping.

I used some additional images to see what kind of results my algorithm would produce on different types of pictures.
 

I was brought in to modify the Footprints workspace and create a front end that tutors could use to input new tutor sessions and record the attendees. In the end, the front end became the primary tool used by the ASC and Footprints was providing nothing more than a database structure for the application. In September 2011, I modified the front end to use an independent set of MySQL tables and Footprints was no longer required.

The application allows an administrator to create locations where tutor sessions can occur. This information can be used later on when reports are generated.

At the beginning of each semester an administrator needs to upload a list of Tutors and the sessions they are qualified to teach.

Then a full listing of all the student body and the courses they are enrolled in is uploaded.

After the tutors and students have been loaded, each individual tutor can log in and create tutor sessions and list the enrollment.

If a student is not enrolled in the course the tutor is teaching, they can add the student manually.

After saving the course, the tutor will receive notification if it was successful or if an error occurred.

The Footprints implementation provided a more robust set of session editing functionality, but the ASC office never used the Footprints interface, so some simple session editing capabilities are available in the front end.

The final piece of the ASC system is a reporting tool which allows a straight data dump by date, or more detailed reports. All the data is exported to a CSV file for Excel. This allows the administrators to manipulate the data in a tool that they are comfortable with and allows for easy distribution of those reports to people who are not familiar with the ASC tutor tracking application.

For the past three years, I have been working on and perfecting a customer queuing system. The original concept was developed as a response to our customers standing in line at the CCIT Support Center. At the beginning of each fall and spring semester, we would have students lined up out the door as our small help desk staff of 5 – 6 people tried to handle all the issues using a first come first serve model. The customers would get upset because they would stand in line forever, even for a simple plot pickup or password reset. Our staff were not being utilized efficiently because each person on the desk needed to have a very broad skill set in order to handle anything the next customer might need, from software issues to plot pickups.

Taking a cue (no pun intended) from the Department of Motor Vehicles, we decided a queuing system where a user would take a number and be helped by the next available agent trained in their specific need was the best solution. At first the queuing system simply accepted the user’s name, a reason for their visit, and asked them to have a seat. We installed chairs and couches, which were really well received by the customers. With the new information, we were able to generate reports on how long users were waiting, being assisted at the desk, and what type of issues they were visiting for.

The reporting information helped us streamline and soon we had express lines opened for requests that took under two minutes and users were flying through the support center. We went from customers waiting in line for an hour or more, to averaging under 10 minutes per customer. The queue system was expanded to include new options like integration with our ticketing system for more accurate ticketing and reporting, multiple desk locations for our satellite help desks around campus, and alerts to supervisors (visual and audible) when waits exceed specified amounts. The queue screen displayed to customers in our waiting area now includes a news crawler section that keeps them entertained and informed while they wait.

The system was developed using a combination of PHP, HTML, CSS, JavaScript (with AJAX), and Java. I have included some screenshots below.

The Sign In Screen can be displayed on one screen horizontally or vertically, or it can be displayed on a series of screens for more complicated sign in processes.

The Queue Screen displays the current queue for customers and some helpful information.

The Agent Screen allows agents to assist customers, complete customers, and interact with their ticketing system tickets.

The Supervisor Screen allows supervisors to monitor the current activity and receive alerts from the queue system.

The Admin Screen allows administrators to create new locations, configure those locations, define reasons for a customer visit, and report on activity.

The game can be run on most any computer if you download an NDS emulator (http://en.wikipedia.org/wiki/Nintendo_DS_emulation). I recommend No$GBA because that is what I tested the game in. Not all the features work on a computer (like sound and saving), but if you actually decide to load the rom on an a Nintendo DS it will work as expected. I’ve included some screenshots and the NDS rom file.

2009_PirateMaze.nds

After taking the photo, I used Photoshop Express on the iPhone to tint the photo blue.

Because I only wanted my skin to be blue, I opened the original picture in Brushes on the bottom layer. Then I dropped the blue tinted picture on the next layer up and carefully erased everything that wasn’t skin to reveal the original image.

Finally, I used the painting tools in Brushes to black out my teeth and add a cute tongue, because Smurfs don’t seem to have teeth.

Click HERE only if you don’t get it.

See the original http://failbook.failblog.org/2011/08/11/funny-facebook-fails-john-from-nebraska/