An image‐based approach for quantitative assessment of uniformity in particle distribution of noise reduction material

• Published in: Microscopy research and technique Vol. 84; no. 9; pp. 1924 - 1935
• Main Authors: Tan, Hua, Huang, Zhang‐Can, Zhu, Si‐Rong, He, Lang, Fang, Xi
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.09.2021; Wiley Subscription Services, Inc
• Subjects: Algorithms; Background noise; Equalization; Histograms; image binarization; Image contrast; Image enhancement; Image processing; Image quality; Morphology; Noise; Noise pollution; Noise reduction; Otsu; Porosity; Sound transmission; Statistical analysis; uniform distribution; uniformity quantification; Visual observation; image processing; Otsu; image binarization; uniformity quantification; uniform distribution
• ISSN: 1059-910X; 1097-0029; 1097-0029
• DOI: 10.1002/jemt.23748

Under the background of noise pollution caused by railway development, noise reduction material is worthy of in‐depth study. A uniform distribution of particles in the material has an important influence on sound absorption property. In this article, the relevant image processing technology is applied to get structure information to quantify the uniformity. The main contributions of this study are: (a) In the preprocessing stage, SEM cross‐sectional image of material is processed by mean filter and histogram equalization. Therefore, the grayscale and the contrast between target and background are enhanced, and a low‐quality image is transformed into a high‐quality one. (b) In the locating stage, local details of the image are considered to discriminate each particle from the whole image. When a global threshold is combined with the local iteration threshold, an improved Otsu algorithm is designed to binarize the image. Through morphology transforming, area filtering, and hole filling, the connected domain of target can be found and particles are located. (c) In the assessing stage, area index, number index and local distance index are established for assessing the uniformity of pore distribution. The experimental results indicate that statistical analysis is consistent with human visual observation. The smaller the porosity is, the better the uniformity is. Compared with some important methods, the effectiveness and efficiency of the proposed approach could be illustrated. An image‐based approach for quantitative assessment of uniformity is proposed. The local iteration Otsu algorithm is applied to binarize the image. The three assessment indexes are established for quantifying the uniformity of particle distribution.