基于KE-PCA法融合眼动数据的黎平县苗侗民族刺绣绣种–意象认知研究
Research on Embroidery Species-Image Cognition of Miao-Dong Ethnic Embroidery in Liping County Based on KE-PCA Method and Eye Movement Data

针对传统服饰体验设计评价中缺乏刺绣的感性意象研究，本文提出一种基于KE-PCA法融合眼动数据的黎平县苗侗民族服饰刺绣绣种–意象认知评价方法。首先，运用KE法提取代表性视觉意象集，使用系统聚类法提取代表性意象词对；其次，使用眼动追踪技术提取代表性绣种工艺，要求专业绣娘制作单一代表性样本实物组成代表性样本图集；接着，获取主观评价和眼动生理数据采用单因素方差分析筛选指标；最后，基于PCA法融合多维指标建立综合评价模型。最终依次计算刺绣绣种–意象评价综合特征值，以此获得绣种–意象排序关系，继而提炼出与用户视觉认知关联的绣种及其权重排序推断出刺绣体验设计的优选方案，并以苗族刺绣制品为例验证其可行性。本文旨在为协助面向用户感性体验的服饰刺绣创新设计和感性服饰体验设计提供民族服饰刺绣的意象选择和参考。
Aiming at the lack of perceptual image research of embroidery in the evaluation of traditional clothing experience design, this paper proposes a cognitive evaluation method based on the KE-PCA method and eye movement data for the embroidery of Miao-Dong ethnic costumes in Liping County. Firstly, we use KE method to extract representative visual image sets, and use systematic clustering method to extract representative image word pairs; secondly, we use eye tracking technology to extract representative embroidery techniques, requiring professional embroidery mothers to produce a single representative sample of physical composition representatives sample atlas; then, we obtain subjective evaluation and eye movement physiological data and use single-factor analysis of variance to screen indicators; finally, we build a comprehensive evaluation model based on the PCA method and fusion of multi-dimensional indicators. Ultimately, the comprehensive feature value of embroidery type-image evaluation is calculated in sequence to obtain the embroidery type-image ranking relationship, and then the embroidery types related to the user’s visual cogni-tion and their weight ranking are inferred to infer the optimal plan of embroidery experience de-sign. Taking Miao embroidery products as an example to verify its feasibility, this article aims to provide the image selection and reference of ethnic costume embroidery for assisting the innova-tive design of costume embroidery and the sensual costume experience design for users’ perceptual experience.