Monday

Today I researched the different kinds of connectedness and the methods of identifying connected components in binary images. I want to implement the more efficient sequential algorithm, since I am writing this code for timing and scheduling purposes. This requires a sort of lookup table which I am unsure how to implement. There are two possible implementations, an array or a linked list, and I am not sure which would be more efficient, so I e-mailed Margrit (she was not in her office) and asked her advice.

Tuesday

Today I implemented the first half of the sequential algorithm for identifying connected components in binary images. Each black pixel is now given a component label which is found by checking its top and left neighbors. If neither of these two pixels has a label, a new label is started for the pixel. If one neighbor has a label, the pixel is given the same label. If both have the same label, the pixel is also given the same label. If the pixels both have a label but they are not the same, the pixel is given the label of the top, earlier checked pixel, and the left pixel's label is entered into an equivalency table. I am not sure how to go about the next step, resolving the equivalency issues. Since I had not heard from Margrit, I implemented the table as an array, but I may have to switch it to a linked list. I also created a new image, which displays each component in a different shade of gray.

Wednesday

Stephen e-mailed us last night and said he made some changes to the Finger Counter and would like us to test it. We decided that it would be really nice if we could run it off our personal laptops so we could ask random people on the street to test it for us. This necessitates us installing Linux, which we both wanted to do this summer, so now we will have help setting up the partition and getting everything working. I went home and got both of our laptops, and then we started backing up our data so that tomorrow we can set up a partition for Linux and then install Red Hat 9. Christen also started running the timing code on the function which creates the binary image. Right now we don't have much to work with, Margrit just wanted us to have something to present at the Sensorium meeting today.

Thursday

We spent all day today trying to get Partition Magic to work. The main problem was nobody had it. The Personal Computing Support Center downstairs said they did, but it turned out to be a burned copy of a demo CD (Why burn a copy of a demo CD? I have no idea.) so it would pretned it was going to re[artition our computers, and then ask us for money, yet the technician doing it for us, since we weren't allowed to do it ourselves, just kept trying and trying.

Friday

Happy Fourth of July!