MakeHuman: A Review of the Modelling Framework

Abstract

MakeHuman is an open source software rarely used in Ergonomic studies. Developed on open source Python code, the program creates realistic appearance 3D virtual human models, primarily focusing on morphing details. An intuitive graphical user interface working with sliders controls input parameters on normalized scales for the main parameters gender, age, muscle mass, weight, height, proportion and ethnicity. These input parameters govern associated output values, which mostly remain normalized. Height and age however are on an interval scale. MakeHuman Blender tools connect the MakeHuman and Blender programs, allowing users to modify a base mesh shape, create clothes, apply static poses or generate animations. In recent research work, MakeHuman was employed mostly to generate sets of virtual subjects. MakeHuman is a design (gaming) oriented, parametric virtual human modelling tool based on templates. A template model is transformed by means of scaling factors, resizing its segments and proportions, to create a set of human bodies compatible with the original base mesh. The template model is divided into ‘areas of influence’, and form factors are calculated to detect contraction or expansion, improving the use of targets in these areas. Fuzzy logic rules are employed in order to process inputs, which are linked directly to membership functions of fuzzy sets. With one morphing target file for each parameters’ extreme values, multifactorial input change is amalgamated into a character, using an inference engine that produces a diversity of human bodies. The study aspires to assess the practicability of using the software in a Human Factors framework.