5th week Updates

During this week, we concluded our algorithm for the field detection. We tested our code using the images in the DRIVE database to see whether our algorithm worked, and we successfully implemented our algorithm, as the results were the ones desired. When crafting this code, it was essential for us to covert the RGB retinal image into HSV to analyze the intensities. It’s easier to detect the outer space instead of the field, so the mask we create in the code will be of the outer space. To find the threshold necessary to detect the outer space took some trial and error. After trying .25 we realized that we were getting too much noise. So, after more trials with more values, we settled on .13 as the best threshold for detecting the outer space. Then, we used the inverse of the outer space mask we created, to give us the inner space, the field.

While this was a busy academic week for us, we still managed to identify more papers for literature review, and these documents are on the topic of optical disk detection, another component of our project.

Database of images used as reference and test: http://www.isi.uu.nl/Research/Databases/DRIVE