Path planning directed motion control of virtual humans in complex environments

Highlights

• Generate realistic motions naturally adapted to complex dynamic environment;

• Apply the optimal path planning to direct motion synthesis of virtual humans;

• SIPP is implemented for planning the optimal path in complex dynamic environment;

• Three types of control anchors are extracted to direct motion field synthesis;

• Both the optimal path and the realistic motions are guaranteed simultaneously.

Abstract

Natural motion synthesis of virtual humans have been studied extensively, however, motion control of virtual characters actively responding to complex dynamic environments is still a challenging task in computer animation. It is a labor and cost intensive animator-driven work to create realistic human motions of character animations in a dynamically varying environment in movies, television and video games. To solve this problem, in this paper we propose a novel approach of motion synthesis that applies the optimal path planning to direct motion synthesis for generating realistic character motions in response to complex dynamic environment. In our framework, SIPP (Safe Interval Path Planning) search is implemented to plan a globally optimal path in complex dynamic environments. Three types of control anchors to motion synthesis are for the first time defined and extracted on the obtained planning path, including turning anchors, height anchors and time anchors. Directed by these control anchors, highly interactive motions of virtual character are synthesized by motion field which produces a wide variety of natural motions and has high control agility to handle complex dynamic environments. Experimental results have proven that our framework is capable of synthesizing motions of virtual humans naturally adapted to the complex dynamic environments which guarantee both the optimal path and the realistic motion simultaneously.

Introduction

Animating and controlling virtual humans realistically in complex dynamic environments are a crucial problem in computer animation, which depends on both the path virtual humans choose and the character motions synthesized. Many efforts have been made on the two aspects of path planning and motion synthesis respectively and great advancements have been achieved. However, it is still a challenging task in computer animation to create realistic and natural motions of character animation actively responding to complex dynamically varying environments.

To solve this problem, a novel comprehensive framework has been proposed in this paper which directs motion synthesis of virtual humans in complex dynamic environment by a global optimal path planned.

For path planning, SIPP algorithm [1] is used as path planner to find a globally optimal path from current location in a complex dynamic environment to another where virtual character want to go in the presence of dynamic obstacles, which adds time as an additional dimension into the search-space explored by the path planner to properly handle moving obstacles and introduces the concept of safe time intervals to greatly reduce the number of states that need to be searched. As the control information for directing motion synthesis of virtual character that drives virtual human follow the path planned naturally and realistically, three types of control anchors are defined, including turning anchors, height anchors and time anchors, and extracted from the path planned.

For motion synthesis, a popular traditional approach is motion graphs, which gains the motion that satisfies user demand by compressing connections among different pieces of motions in database and searching this graph. Although this is conceptually intuitive and broadly applicable, graph-based methods lack the adequate flexibility and controllability to synthesize agile reactivity of the animated character to the dynamically changing environment due to the sparse and finite states in the graph. In our framework, motion fields [2] approach is used as motion synthesizer, which organizes motion dataset into a high-dimensional generalization of a vector field of state space, then generates an animation by freely flowing through the motion field in response to interactive controls. Since each state in motion field has a set of candidate actions and Reinforcement Learning is adopted to find an action that leads to a desirable control, these enable real-time controlled highly responsive motion synthesis for motion field instead of waiting for pre-determined transition points for motion graph. So motion field is able to control character motion at arbitrary state in a continuous state space for motion synthesis, which allows highly agile motion controls with control anchors for adapting to dynamically varying environment and guaranteeing the synthesized character motion well-fitting to the global optimal path planned.