Scholarlink

<p>Visual balance is one of the most important factors affecting the efficiency of map information transmission and visual aesthetics. Although the existing balance calculation models can provide better measurements and evaluations for traditional maps, they have little applicability to WeMap with personalized element configurations. In particular, the problems of difficult extraction of map configuration elements, single visual feature factor construction, and subjective visual balance degree discrimination pose great challenges to the visual balance calculation of WeMap. As such, a visual balance calculation method for WeMaps combining fuzzy clustering and multi-factor driving is proposed in this paper. First, each WeMap is subjected to morphological reconstruction and affiliation matrix division to optimize the distribution characteristics of the image, determine the attribution of each element, and realize the classification and extraction of map configuration elements based on robust fuzzy C-mean clustering. Then, we introduce three visual perception factors, brightness, contrast, and saliency, which focus on the weight, detail, and focus of the image, respectively. Combined with the positional factors affecting the layout and distribution of the elements, which together constitute the four components of visual density, the W-VBC computational model is then constructed. Finally, based on the expert evaluation knowledge and image computational features, we build the WeMap visual balance discriminative model and introduce multi-factor Bayesian decision-making to effectively divide the visual balance area and avoid the problem of non-linearly differentiable determination results. The experimental results show that the proposed W-VBC model exhibits good performance on a dataset consisting of 1040 WeMaps, with an accuracy of 70.15% for balance judged as balanced, 71.56% for imbalance judged as unbalanced, 6.42% for balance misjudged as unbalanced, and 2.98% for imbalance misjudged as balanced. The calculation results of this method are consistent with the expert mapping experience, which reflects the better simulation performance of human visual cognition.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A visual balance calculation method for wemaps combining fuzzy clustering and multi-factor driving

  • 期刊论文

  • 发表日期：

    2025年8月9日

查看全文
  • Rongjuan Ma,
  • Haowen Yan,
  • Xiaomin Lu,
  • Zhaoyang Hou,
  • Xiaolong Wang,
  • Pengbo Li,
  • Tianshu Chu,
  • Fukang Mao

摘要

Visual balance is one of the most important factors affecting the efficiency of map information transmission and visual aesthetics. Although the existing balance calculation models can provide better measurements and evaluations for traditional maps, they have little applicability to WeMap with personalized element configurations. In particular, the problems of difficult extraction of map configuration elements, single visual feature factor construction, and subjective visual balance degree discrimination pose great challenges to the visual balance calculation of WeMap. As such, a visual balance calculation method for WeMaps combining fuzzy clustering and multi-factor driving is proposed in this paper. First, each WeMap is subjected to morphological reconstruction and affiliation matrix division to optimize the distribution characteristics of the image, determine the attribution of each element, and realize the classification and extraction of map configuration elements based on robust fuzzy C-mean clustering. Then, we introduce three visual perception factors, brightness, contrast, and saliency, which focus on the weight, detail, and focus of the image, respectively. Combined with the positional factors affecting the layout and distribution of the elements, which together constitute the four components of visual density, the W-VBC computational model is then constructed. Finally, based on the expert evaluation knowledge and image computational features, we build the WeMap visual balance discriminative model and introduce multi-factor Bayesian decision-making to effectively divide the visual balance area and avoid the problem of non-linearly differentiable determination results. The experimental results show that the proposed W-VBC model exhibits good performance on a dataset consisting of 1040 WeMaps, with an accuracy of 70.15% for balance judged as balanced, 71.56% for imbalance judged as unbalanced, 6.42% for balance misjudged as unbalanced, and 2.98% for imbalance misjudged as balanced. The calculation results of this method are consistent with the expert mapping experience, which reflects the better simulation performance of human visual cognition.