This book focuses on utilizing visual information for real-time reconstruction and understanding of scenes to achieve active interaction functions. Starting from areas such as robot-scene modeling, understanding, navigation, and interaction, it comprehensively explains the relevant technologies and specific practices involved. The book revolves around the theme of scene modeling and robot perception, establishing a knowledge system that integrates multiple disciplines, primarily targeting senior undergraduate students, graduate students, and related technical personnel interested in computer graphics, computer vision, and robotics. Combining theory with practice, this book contains the most basic algorithms, accompanied by detailed code or pseudo-code resources, as well as special content such as “chapter knowledge points”, aiming to provide effective references for personnel working in this field.