In this monograph, both traditional and modern object segmentation approaches are investigated, comparing their strengths, weaknesses, and utilities. The main focus is on the deep learning-based techniques for the two most widely solved segmentation tasks: Semantic Segmentation and Instance Segmentation.