A hybrid shape-based image clustering using time-series analysis
- PDF / 2,107,876 Bytes
- 16 Pages / 439.37 x 666.142 pts Page_size
- 91 Downloads / 291 Views
A hybrid shape-based image clustering using time-series analysis Atreyee Mondal 1 & Nilanjan Dey 1 & Simon Fong 2 & Amira S. Ashour 3 Received: 19 March 2020 / Revised: 11 July 2020 / Accepted: 28 August 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Clustering of different shapes of the same object has an inordinate impact on various domains, including biometrics, medical science, biomedical signal analysis, and forecasting, for the analysis of huge volume of data into different groups. In this work, we present a novel shape-based image clustering approach using timeseries analysis, to guarantee the robustness over the conventional clustering techniques. To evaluate the performance of the proposed procedure, we employed a dataset consists of various real-world irregular shaped objects. The shapes of different objects are first extracted from the entire dataset based on similar pattern using mean structural similarity index. Furthermore, we performed radical scan on the extracted shapes for converting them to one-dimensional (1D) time-series data. Finally, the time series are clustered to form subgroups using hierarchical divisive clustering approach with average linkage, and Pearson as distance metrics. A comparative study with other conventional distance metrices was also conducted. The results established the superiority of using Pearson correlation measure, which provided the maximum F1-score with exact number of shapes under a sub-cluster, while the corresponding outcomes of other approaches results in a poor and inappropriate clustering. Keywords Shape clustering . Image to time-series conversion . Shape similarity . Hierarchical clustering . Pearson correlation distance metric
* Amira S. Ashour [email protected]; amira.salah@f–eng.tanta.edu.eg
1
Department of Information Technology, Techno India College of Technology, Kolkata, West Bengal, India
2
Department of Computer and Information Science, University of Macau, Taipa, Macau SAR
3
Department of Electronics and Electrical Communications Engineering, Faculty of Engineering, Tanta University, Tanta 31527, Egypt
Multimedia Tools and Applications
1 Introduction Clustering is a significant unsupervised learning algorithm. Time-series clustering has a wide range of applications in the field of medical science, and business [7]. The conventional clustering methods are based on partitioning algorithms, such as k-means, k-medoids, model-based clustering, and hierarchical clustering. Hierarchical clustering methods results good accuracy with time-series data, while the other clustering algorithms deal with static data [30]. In addition, the shape-based clustering facilitates wide application domains such as cytology, video data analysis, healthcare where different types of cells need to be clustered without human supervision for detection of diseases, and documentation of genotypes [22]. Shape clustering can be used in grouping objects in a surveillance system, grouping of species, pathological cell clustering, an
Data Loading...
