End-to-end concrete appearance analysis based on pixel-wise semantic segmentation and CIE Lab
End-to-end concrete appearance analysis based on pixel-wise semantic segmentation and CIE Lab
复制标题
基于像素语义分割和 CIE Lab 的端到端混凝土外观分析
DOI：
10.1016/j.cemconres.2022.106926
复制
发表时间：
2022
影响因子：
11.4
通讯作者：
Xinyu Qi
中科院分区：
文献类型：
--
作者：
Zhexin Hao;Xinyu Qi
传统的专家现场验收方式已经不能满足现代混凝土行业智能化的需要。计算机视觉领域的前沿成果极大地加速了混凝土表面相关工程的自动化进程。我们提出了一种基于像素级语义分割的端到端具体外观多重识别方法以及基于CIE Lab的相应的定量色差分析方法。所提出的基于全卷积网络（FCN）的网络的目标函数在经验风险和结构风险方面进行了优化，并降低了复杂性以提高其泛化性能。比较不同的预设基础 CNN 后，最佳结果显示 98.43% 像素精度 (PA)、91.33% 平均像素精度 (MPA)、83.89% 平均交集 (MIoU) 和 96.95% 频率加权并集交集 (FWIoU)。此外，还探讨了不同光强度和表面水分含量下的类内变化，以证明该方法的稳健性。
The traditional expert-on-site acceptance method has been unable to meet the need of intellectualization of modern concrete industry. The cutting-edge achievements in the field of computer vision have immensely accelerated the automation of concrete surface-related engineering. We propose an end-to-end concrete appearance multiple identification method based on pixel-wise semantic segmentation and a corresponding method for quantitative chromatic aberration analysis based on CIE Lab. The objective function of the proposed Fully Convolutional Network (FCN)-based network was optimized in terms of both empirical and structural risks, and the complexity was reduced to improve its generalization performance. After comparing different preposed base CNNs, the best results show 98.43 % pixel accuracy (PA), 91.33 % mean pixel accuracy (MPA), 83.89 % mean intersection over union (MIoU), and 96.95 % frequency weighted intersection over union (FWIoU). In addition, intra-class variation was explored under different light intensity and surface moisture content to demonstrate the robustness of the proposed method.
影响因子：
11.4
作者：
Guo, Pengwei;Meng, Weina;Bao, Yi
通讯作者：
Bao, Yi