Three-dimensional video (3D video) has seen significant advancements in fields such as film, television, gaming, and virtual reality, enhancing visual experiences by adding depth perception. However, the data demands of 3D video far exceed those of traditional 2D video, posing challenges in storage and transmission. This paper introduces three types of 3D video: stereoscopic video, multi-view video, and free-viewpoint video, and then discusses the evolution of compression techniques for these formats. From early dual-stream compression to mid-term developments like disparity compensation and multi-view video coding (MVC), and finally to modern methods such as multi-view plus depth (MVD) and the latest applications of deep learning, 3D video compression technologies have evolved to meet increasing.