In the last decade, a large variety of algorithms have been developed for use in object tracking, environment mapping, and object classification. It is often difficult for beginners to fully predict the constraints that multirotors place on machine vision algorithms. The purpose of this paper is to explain some of the types of algorithms that can be applied to these aerial systems, why the constraints for these algorithms exist, and what could be done to mitigate them. This paper provides a summary of the processes involved in a popular filter-based tracking algorithm called MOSSE (Minimum Output Sum of Squared Error) and a particular implementation of SLAM (Simultaneous Localization and Mapping) called LSD SLAM.