Work

Perbandingan Jarak Euclidean, Cityblock, Minkowski, Canberra, Dan Chebyshev Dalam Sistem Temu Kembali Citra Batik

Year: 2024

Type: article

Abstract: Batik is a highly valuable cultural heritage in Indonesia, showcasing a rich diversity of motifs with deep meaning and aesthetics. To enhance the accessibility and utilization of batik collections, an efficient image retrieval system is essential. This study compares distance measurement methods in a batik image retrieval system: Euclidean, Cityblock, Minkowski, Canberra, and Chebyshev, using a combination of color and texture features. The dataset comprises 50 types of batik images. The results show that the Cityblock method achieves the highest Mean Average Precision (MAP) of 97.71, followed by Canberra with MAP 96.87. The Euclidean method also performs well with a MAP of 94.56, while Minkowski and Chebyshev have lower MAP values of 92.93 and 90.89, respectively. Chebyshev experiences the largest MAP drop when images are rotated (5.98), while Cityblock demonstrates the best resistance to rotation with the smallest MAP drop (1.51). This research successfully developed a Content-Based Image Retrieval (CBIR) system with a GUI in MATLAB and suggests integrating the latest image processing and machine learning techniques for further enhancement.

Source: Jurnal Informatika dan Teknik Elektro Terapan

DOI: 10.23960/jitet.v12i3s1.5324

Authors: Mutmainnah Muchtar, Noorhasanah Zainuddin, Adha Mashur Sajiah, Nurfitria Ningsi, Yuwanda Purnamasari Pasrun

Institutions: Universitas Sembilanbelas November Kolaka, Universitas Halu Oleo

Language: English

Topic: Image Retrieval and Classification Techniques

Subfield: Computer Vision and Pattern Recognition

Field: Computer Science

Domain: Physical Sciences

SDG: Sustainable cities and communities

Open Access status: diamond

OpenAlex ID: W4403347668

FWCI:

Cites:

Cited by:

Related to: 10