Curved path human locomotion on uneven terrain
Animation of human figures is a fascinating and complex problem, and with applications in virtual reality, military training, and entertainment, it is also one with great value. One of the most important actions to support is walking. Unfortunately, continued use of scripted animation is becoming untenable: the process of creating animation data becomes a bottleneck. To address this issue, we have investigated data-driven, procedural models for the kinematic animation of human walking. We utilize a new motion data representation, the sagittal elevation angles, and describe a procedural algorithm to generate gait animation given sagittal elevation angle datasets. The use of motion data allows for realistic walking, while procedural animation yields the ability to generate different animation, including curved locomotion on uneven terrain, without requiring new motion data for each variation. We also describe our solution to the inverse motion interpolation problem. This allows our system to operate automatically, with no user interaction needed to supply necessary interpolation parameters. Finally, we also examine the problem of motion dataset generation, with the aim of reducing the reliance on motion data by allowing our system to generate it. Our system generates gait for different sized figures in real-time, and allows for curved path locomotion on uneven terrain, with flexibility for stylistic variation.
Sun, Harold Chi-hao, "Curved path human locomotion on uneven terrain" (2000). Dissertations available from ProQuest. AAI9989658.