"Human Motion Sensing and Recognition - A Fuzzy Qualitative Approach" is oriented toward students, business and researchers involved in research and applying fuzzy logic principles in every day motion tracking and sensing applications. Many of these approaches are new and evolving.
Human Motion Sensing and Recognition
A Fuzzy Qualitative Approach
H. Liu , Z. Ju , X Ji , CS. Chan and M. Khoury
2017 | 281 pages | ISBN: 9783662536902
Advanced technologies are now being used to track and locate humans, as well as being able to recognize faces and movement characteristics. In this book, "The topics covered include human motion recognition in 2D and 3D, hand-motion analysis with contact sensors, and vision-based view-invariant motion recognition, especially from the perspective of Fuzzy Qualitative techniques."
There is a significant amount of ongoing research in this area along with software development. The book is arranged to include both theory and applications involving fuzzy logic. Opening with topics related to vision-based sensing, wearable sensing and multimodal sensing, various techniques for sensing are described. Recognition approaches including probabilistic graphical models, support vector machines, artifical neural networks and fuzzy approaches. The text also includes mathematical formulas for calculating and understanding these applications.
Since the future is likely to include robotics, information about the multidisciplinary research efforts in this area are also included. Robotics play a unique role since nnot only are their location of signficance, but also recognizing them will be a significant and compelling challenge. Intelligent connection issues and communication approaches will be needed to harness their full potential. In a similar manner, the authors investigate human motion sensing approaches. Readers here will find a significant amount og mathematical heory describing the fuzzy logic approaches and methods used.
Human motion is intriguing due to the detailed approaches and ties to anatomy that are being considered, particularly in terms of natural movements and observing them for various parts of the body. While much of this might seem straight forward to readers, it should be remembered that mathematical formulas are being used to effectively represent natural movement forms and their reactions in 3D spaces.
Fuzzy Gaussian Mixture Models are described in depth. These models attempt to weight membership into classifcations of movement variables using statistical techniques. These cluster methods are described and in detail. Hand models have been given significant attention, since it is anticipated that hand movements will perform most or many of common robotic-like tasks. These models aim to include multi-fingered devices capable of actions and movement like human hands. At the same time, it should be recognized that hand movements must link with capable sensing technologies that can interpret force, direction, movement and so on.
Accordingly, multi-sensor technologies capable of meeting these challenges are also discussed in detail. IN the case of movements and gestures, these technologie will often deploy visualization technologies coupled with image analysis. These software need to be able to understand gestures and movements while also recognizing and interpreting them. The authors include a significant number of images together with text that allows readers to see and understand the meanings of the written text.
All in all, "Human Motion Sensing and Recognition - A Fuzzy Qualitative Approach" provides a comprehensive understanding of current issues in motion traching and sensing. It is based upon a large number of research efforts that are often being implemented in their initial form. The authors help students and readers to understand the mathematical approaches being used to advance these technologies.