The book focuses on a new research direction of image understanding——spatio-temporal behavior understanding, mainly involving detection and recognition technologies at all levels from action primitives, actions, activities, events to behaviors. This book focuses mainly on introducing the principles of these technologies, analyzing relevant methods combined with research findings, and summarizing recent research advances.

In addition to the introduction, the book consists of seven chapters, covering respectively the detection of spatial-temporal interest points and human key points, motion trajectory analysis of humans and objects, classification and recognition of human actions, modeling and recognition of human activities, behavioral analysis and reasoning integrating various cues, detection and judgment of abnormal events, as well as detection and interpretation of human-object interactions.

This book can serve as a professional textbook or teaching reference for graduate students in disciplines such as Signal and Information Processing, Communication and Information System, Electronic and Communication Engineering, and Pattern Recognition and Intelligent Systems. It also supports researchers in computer science and technology, information and communication engineering, electronic science and technology, measurement and control technology and instruments, robotics and automation, biomedical engineering, optics, electronic medical device development, remote sensing, surveying and mapping, and military reconnaissance in conducting relevant research.

ZHANG Yujin is a Professor at Tsinghua University, and his research interests include Image Engineering, which covers image processing, image analysis, image understanding and their technical applications, as well as related teaching theories and methodologies.