高级检索

基于视觉主观评价的绿色工地评价应用研究

Research on the Application of Green Construction Site Evaluation Based on Visual Subjective Evaluation

  • 摘要: 为解决传统绿色工地评价方法忽略主观感受、难以直观呈现的问题,选取安全性、美观性、高效性作为评价维度展开绿色工地视觉主观评价的研究,构建了基于TrueSkill算法的评分网站系统以探索视觉主观评价方法的应用思路。所述方法是选取106名受访者对在两个实际工地收集到的1500张图像,从上述三种评价维度进行评价,并利用Trueskill算法进行匹配计分。平均每名受访者完成16组评价测试,获得共4086份评价数据,通过分析Skill值散点图、Skill值频率分布曲线及图像分数变化,证实了基于Trueskill算法的视觉主观评价方法在绿色工地评价的场景下具有充分可信性。此外,上述研究过程中收集的感兴趣区域(Region of Interest,ROI)数据可作为计算机视觉在绿色工地评价应用研究的数据基础,对计算机视觉在绿色工地实时监管方面的研究有参考意义。

     

    Abstract: To address the issues of traditional green construction site evaluation methods that overlook subjective feelings and are difficult to present intuitively, the green construction site evaluation research was developed while safety, aesthetics, and efficiency were taken as evaluation dimensions, and a scoring website system based on the TrueSkill algorithm was then constructed to explore the application approach of visual subjective evaluation methods. The mentioned method was conducted by scoring 1500 images collected from two real construction sites by 106 respondents in terms of three evaluation dimensions above. On average, 16 sets of evaluation tests were completed by each subject, yielding a total of 4,086 evaluation data and the experimental data analysis through figure scores, scatter plots and frequency curves on Skill value, shows that the above research method based on visual subjective evaluation reached sufficient credibility. In addition, the ROI (Region of Interest) data collected in the above study can serve as the data foundation for the application research of computer vision in green construction site evaluation, and has reference significance for the study on computer vision in real-time supervision of green construction sites.

     

/

返回文章
返回