An analysis of a kernel of corn is essential to determining its worth for selective breeding. The embryo found in a kernel of corn is what is used to produce oils, and therefore maximizing its size would make a farmer's yield more efficient. The current method of recording the area of the corn's embryo, however, is to take measurements by hand. This project seeks to provide a solution by developing a program that can recognize and mark the embryo on a picture of a kernel of corn without needing guidance. The program will be taught to recognize patterns in thousands of images which have already been marked, until it can successfully apply those patterns to mark the embryo on an unlabeled kernel. A user interface will then be developed to allow researchers a simpler method of employing the program, simply uploading images to it and running the application. When complete, this project will expedite the kernel marking process significantly, allowing researchers more time to do more extensive and in-depth analysis on the data collected. This could likely result in faster improvements to embryo size as well as more impressive end results.