Week 8 Progress Update

During the past week, we continued working on the blood vessel detection coding. Through further investigation of our code, we realized our next step was to implement an adaptive local and global threshold. For example, we used the Otsu method on the retinal image to find the global threshold to separate the two groups. We quickly discovered that when applied to an image with an uneven background, it isn’t optimal for detection. So, in the following weeks we will continue to apply local and global thresholds in attempts to generate a normalized image of a retinal image that bestly distinguishes the blood vessels.