Gait Synthesis of Abnormal Gaits in Canines

Abstract

When attempting to introduce new subjects or characters to a motion database, generating or capturing motions is particularly cumbersome, especially if the subject has some medical issue. This thesis describes techniques in which motion capture analysis is used to isolate the components of canine gait that are the direct result of poor health. These components are then synthesized with the healthy motion of a new dog to create the appearance of the unhealthy motion in that breed. Given two captured motions, one healthy A and one unhealthy A?, the goal of this work is to determine a mathematical method for finding the difference between A and A? and applying this difference to a third motion, B, to generate a new motion, B?, that carries the characteristics shown in A?. This motion analogy method is successfully implemented and validated using two case studies. This technique has applications in diagnosing the impact of treatments in medical studies and the development of training for veterinarians as it allows for multiple variations of motion to be applied to a new subject.