| The detection and tracking of human motion is key to biomechanics, and attach more and more importance recent few years. Research in this domain involves wide application including physical science, medicine, bionic machine and robot. Especially in the tendency of anthropomorphize and aged society, human motion study not only make human life more safe and efficient but also application future and benefits to the society.This paper presents a multi-dimension video system, which could analyze and deal with the image sequence gathered from three fixed CCD camera, recognized and keep tracking of marker motion. Then we apply global optimization (GO) method in order to reduce skin movement artifacts that imply relative movement between markers and bones. And set up a human kinematics model and design an algorithm based on the RBA theory to study the human activity of daily life.Problems about the principle and method of the measurements, camera calibration, acquisition of images, filter and smooth of images is discussed at first when use a video-based analysis system to study human movement.To study the human activity of daily life is important to the design of recover and unbarried machine, and to make the study of human motion more practice meaning. This paper presents the collection of 268 human activities in daily life, and after shooting the process and analyzed them, we got 16 basic motions, and detecting them with the video system to get the parameters of human motion. The result shows that the system has better precision and the advantage of fast. Finally, we put the result in use of the ability evaluation of the elder and the handicapped.Finally the human motion is simulated with the Lifemod. The process is to build a human body model and created the contact forces between the body segments and environment. Models may be animated and dynamic plots displays simultaneously to observe the internal reactions of the model and the effects on the external environment.
|
PDF Full Text Request