Improving Image Retrieval Performance with SCS and MCS Clustering Techniques

Ulul Fikri; Rahmat Prakoso; Yufis Azhar

doi:10.23887/janapati.v13i2.75643

Authors

Ulul Fikri Universitas Muhammadiyah Malang
Rahmat Prakoso Universitas Muhammadiyah Malang
Yufis Azhar Universitas Muhammadiyah Malang

DOI:

https://doi.org/10.23887/janapati.v13i2.75643

Keywords:

image retrieval, Corel, MTCD, clustering, K-Medoids, DBSCAN

Abstract

This paper presents two methods, Single Cluster Search (SCS) and Multiple Cluster Search (MCS), aimed at enhancing image retrieval performance on the Corel1k, Corel5k, and Corel10k datasets, which has a wide variation of images. The Multi Texton Co-Occurrence Descriptor (MTCD) method is used for feature extraction, and the K-Medoids and DBSCAN methods are used for dataset clustering. The clusters are then ranked based on the distance of their medoids to the query image. The most relevant images are retrieved from the highest-ranking clusters. SCS selects the cluster with the highest ranking as the search area and expands the search area to the next ranking cluster if the number of images is less than 6, which is the desired number of retrieval results. MCS merges several clusters with the highest ranking and combines clusters as the search area. Both methods are evaluated using several metrics, such as AP, MRR, and retrieval time. The results are also compared with the original method, which does not use clustering (the query image and the dataset are only extracted with MTCD, and their distance is calculated). The findings indicate that both methods improve the retrieval time. In Corel1k, the SCS method reduces the time complexity by 0.001s, while the MCS method, although not surpassing the original method, still shows potential. In Corel5k, both methods reduce the time complexity by 0.052s in the SCS method and 0.015s in the MCS method. In Corel10k, both methods reduce the time complexity by 0.122s in the SCS method and 0.058s in the MCS method, compared to the original method. These results have practical implications for improving image retrieval efficiency. The paper discusses the reasons behind these results and suggests possible directions for future research.

References

S. R. Dubey and S. Mukherjee, “LDOP: local directional order pattern for robust face retrieval,” Multimed. Tools Appl., vol. 79, no. 9–10, pp. 6363–6382, 2020, doi: 10.1007/s11042-019-08370-x.

R. Ashraf, M. Ahmed, U. Ahmad, M. A. Habib, S. Jabbar, and K. Naseer, “MDCBIR-MF: multimedia data for content-based image retrieval by using multiple features,” Multimed. Tools Appl., vol. 79, no. 13–14, pp. 8553–8579, 2020, doi: 10.1007/s11042-018-5961-1.

A. E. Minarno, F. D. S. Sumadi, Y. Munarko, W. Y. Alviansyah and Y. Azhar, "Image Retrieval using Multi Texton Co-occurrence Descriptor and Discrete Wavelet Transform," 2020 8th International Conference on Information and Communication Technology (ICoICT), Yogyakarta, Indonesia, 2020, pp. 1-5, doi: 10.1109/ICoICT49345.2020.9166361.

A. Zhong et al., “Deep metric learning-based image retrieval system for chest radiograph and its clinical applications in COVID-19,” Med. Image Anal., vol. 70, p. 101993, 2021, doi: 10.1016/j.media.2021.101993.

S. Lee, D. Kim, and B. Han, “COSMO: Content-Style Modulation for Image Retrieval with Text Feedback,” Proc. IEEE Comput. Soc. Conf. Comput. Vis. Pattern Recognit., pp. 802–812, 2021, doi: 10.1109/CVPR46437.2021.00086.

X. Li, J. Yang, and J. Ma, “Recent developments of content-based image retrieval (CBIR),” Neurocomputing, vol. 452, pp. 675–689, 2021, doi: 10.1016/j.neucom.2020.07.139.

K. P. Sinaga and M. S. Yang, “Unsupervised K-means clustering algorithm,” IEEE Access, vol. 8, pp. 80716–80727, 2020, doi: 10.1109/ACCESS.2020.2988796.

Z. S. Younus et al., “Content-based image retrieval using PSO and k-means clustering algorithm,” Arab. J. Geosci., vol. 8, no. 8, pp. 6211–6224, 2015, doi: 10.1007/s12517-014-1584-7.

A. E. Minarno and N. Suciati, “Image retrieval using multi texton co-occurrence descriptor,” J. Theor. Appl. Inf. Technol., vol. 67, no. 1, pp. 103–110, 2014.

A. Joseph, E. S. Rex, S. Christopher, and J. Jose, “Content-based image retrieval using hybrid k-means moth flame optimization algorithm,” Arab. J. Geosci., vol. 14, no. 8, 2021, doi: 10.1007/s12517-021-06990-y.

G. H. Liu, L. Zhang, Y. K. Hou, Z. Y. Li, and J. Y. Yang, “Image retrieval based on multi-texton histogram,” Pattern Recognit., vol. 43, no. 7, pp. 2380–2389, 2010, doi: 10.1016/j.patcog.2010.02.012.

Y. Azhar, A. E. Minarno, Y. Munarko, and Z. Ibrahim, “Image Retrieval Based on Texton Frequency-Inverse Image Frequency,” Kinet. Game Technol. Inf. Syst. Comput. Network, Comput. Electron. Control, vol. 5, pp. 87–94, 2020, doi: 10.22219/kinetik.v5i2.1026.

G. Ravivarma, K. Gavaskar, D. Malathi, K. G. Asha, B. Ashok, and S. Aarthi, “Implementation of Sobel operator based image edge detection on FPGA,” Mater. Today Proc., vol. 45, no. xxxx, pp. 2401–2407, 2021, doi: 10.1016/j.matpr.2020.10.825.

N. Varish et al., “Image Retrieval Scheme Using Quantized Bins of Color Image Components and Adaptive Tetrolet Transform,” IEEE Access, vol. 8, pp. 117639–117665, 2020, doi: 10.1109/ACCESS.2020.3003911.

A. E. Minarno, Y. Munarko, A. Kurniawardhani, and F. Bimantoro, “Classification of Texture Using Multi Texton Histogram and Probabilistic Neural Network,” IOP Conf. Ser. Mater. Sci. Eng., vol. 105, no. 1, pp. 0–7, 2016, doi: 10.1088/1757-899X/105/1/012022.

A. E. Minarno, Y. Azhar, F. D. Setiawan Sumadi, and Y. Munarko, “A Robust Batik Image Classification using Multi Texton Co-Occurrence Descriptor and Support Vector Machine,” 2020 3rd Int. Conf. Intell. Auton. Syst. ICoIAS 2020, pp. 51–55, 2020, doi: 10.1109/ICoIAS49312.2020.9081833.

J. H. Dewan and S. D. Thepade, “Image Retrieval using Weighted Fusion of GLCM and TSBTC Features,” 2021 6th Int. Conf. Converg. Technol. I2CT 2021, pp. 1–7, 2021, doi: 10.1109/I2CT51068.2021.9418131.

D. Deng, “DBSCAN Clustering Algorithm Based on Density,” Proc. - 2020 7th Int. Forum Electr. Eng. Autom. IFEEA 2020, pp. 949–953, 2020, doi: 10.1109/IFEEA51475.2020.00199.

M. Li, X. Bi, L. Wang, and X. Han, “A method of two-stage clustering learning based on improved DBSCAN and density peak algorithm,” Comput. Commun., vol. 167, no. November 2020, pp. 75–84, 2021, doi: 10.1016/j.comcom.2020.12.019.

S. Y. Tan, M. F. Akbar, N. H. M. M. Shrifan, G. N. Jawad, M. Nadhir, and A. Wahab, “Clustering Algorithm-Based Microwave Nondestructive Testing,” 2022.

E. H. S. Atmaja, “Implementation of k-Medoids Clustering Algorithm to Cluster Crime Patterns in Yogyakarta,” Int. J. Appl. Sci. Smart Technol., vol. 1, no. 1, pp. 33–44, 2019, doi: 10.24071/ijasst.v1i1.1859.

E. Herman, K. E. Zsido, and V. Fenyves, “Cluster Analysis with K-Mean versus K-Medoid in Financial Performance Evaluation,” Appl. Sci., vol. 12, no. 16, 2022, doi: 10.3390/app12167985.

Y. Wu, K. Song, F. Zhao, J. Chen, and H. Ma, “Distribution Aligned Feature Clustering for Zero-Shot Sketch-Based Image Retrieval,” 2023, [Online]. Available: http://arxiv.org/abs/2301.06685