Harmonised Shape Grammar in Design Practice

The aim of this thesis is to address the contextual and harmony issues in shape grammar (SG) by applying knowledge from the field of natural language processing (NLP). Currently shape grammars are designed for static models (Ilčík et al., 2010), limited domain (Chau et al., 2004), time-consuming process (Halatsch, 2008), high user skills (Lee and Tang, 2009), and cannot guarantee aesthetic results (Huang et al., 2009). The current approaches to shape grammar produce infinite design and often meaningless shapes. This thesis addresses this problem by proposing a harmonised shape grammar framework which involves applying five levels of analysis namely morphological, lexical, syntactic, semantic, and pragmatic levels to enhance the overall design process. In satisfying these semantically well-formed and pragmatically well-formed shapes, the generated shapes can be contextual and harmonious. The semantic analysis level focuses on the character’s anatomy, body function, and habitat in order to produce meaningful design whereas the pragmatic level achieves harmony in design by selecting relevant character’s attributes, characteristics, and behaviour. In order to test the framework, this research applies the five natural language processing levels to a set of 3D humanoid characters. To validate this framework, a set of criteria related to aesthetic requisites has been applied to generate humanoid characters; these include the principles of design (i.e. contrast, emphasis, balance, unity, pattern, and rhythm) and aspects of human perception in design (i.e. visceral, behavioural and reflective). The framework has ensured that the interrelationships between each design part are mutually beneficial and all elements of the humanoid characters are combined to accentuate their similarities and bind the picture parts into a whole.