Visualization of Human Motion by Placement of Key Poses for Motion with Little Positional Change

This study proposes a method for effectively visualizing 3D motion data by optimizing key pose placement. Visualization using still images is commonly employed in research and application fields to concisely capture the characteristics of motion. However, it can be challenging to differentiate between individual poses and visualize the continuous flow of motion. To address this, the study proposes a method for arranging key poses based on the original motion data and adjusting their placement to avoid overlapping. The direction of motion is considered when arranging the poses, ensuring an accurate representation of the continuous flow of motion. This method aims to enhance the accuracy of visualization and improve the efficiency of motion data analysis and selection. Overall, the study seeks to develop a method that enables clear understanding and accurate visualization of motions with small positional changes.