Fundamentals of Optical Flow

This chapter begins by reviewing several commonly used variational optical flow estimation methods and various state-of-the-art robust optical flow estimation techniques. We explore the intricacies of these methods, with a focus on how they address core challenges such as handling motion discontinuities and environmental variations. Additionally, we analyze the relationship between optical flow estimation and motion segmentation, which is key in applications like object tracking, autonomous systems, and video analysis. Several related works are reviewed to illustrate the progression of the field. Furthermore, this chapter delves into the evaluation measures commonly applied to gauge the performance of optical flow methods, such as angular errors and endpoint errors. Lastly, we discuss some of the limitations inherent in the current state of research, identifying opportunities for improvement in future works.