Lidiawaty, Berlian Rahmy (2020) Analisis Image Retrieval Hasil Surf Pada Marketplace Di Indonesia Untuk Mengenali Nama Motif Batik Solo Berdasarkan Tekstur. Masters thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
09211750054011-Master_Thesis.pdf Download (28MB) | Preview |
Abstract
Setiap motif pada batik Solo memiliki makna tersendiri yang mempengaruhi penggunaannya, ada motif untuk pesta pernikahan atau motif untuk acara pemakaman. Kurangnya pemahaman tentang nama motif batik Solo dapat mempengaruhi penjual di marketplace online di Indonesia dalam memberi nama produk batik Solo yang dijual. Oleh karena itu, penelitian ini memiliki tujuan untuk menilai tingkat akurasi penamaan motif batik Solo oleh penjual di marketplace online Indonesia dengan menemu-kembalikan (image retrieval) citra di hasil pencarian markeplace dengan citra di dataset, sesuai dengan kata kunci nama batik Solo. Proses image retrieval antara citra datates dengan citra dataset akan dilakukan dengan mengenali ciri tekstur dari citra. Metode yang digunakan untuk image retrieval adalah Speeded Up Robust Feature (SURF) dan Brute Force (BF). Terdapat tiga hasil pada penelitian ini. Pertama, persentase ketepatan (PK) penjual pada empat marketplace online dalam menamai produk batik Solo sesuai nama motif batik Solonya sebesar 82,13%. Kedua, nilai akurasi dari metode SURF dan BF dalam menemukan kembali citra untuk mengenali nama motif batik Solo sebesar 51,32%. Ketiga dapat diketahui bahwa marketplace dengan PK tertinggi adalah Lazada dengan nilai 95,42%. Sedangkan marketplace dengan akurasi tertinggi dalam menemubalikan citra dengan metode SURF dan BF adalah Bukalapak dengan nilai akurasi sebesar 74,76%.
===============================================================================================================================
Every pattern in batik Solo has its own meaning that affects its use. There are patterns for specific cultural ceremony, such as wedding or funeral. Therefore, the seller of batik Solo’s product need to be able to give the the right title as the name of their product in Indonesia’s online marketplace, thus the customer will not misswear batik in the wrong event. Based on the importance of naming batik Solo’s product, this study aims to asses the accuracy of batik Solo pattern naming by sellers in Indonesian online marketplace. First it needed to input batik Solo’s pattern name to the marketplace search bar. The second, the study collected the image search results to check wether the seller give the right name of their batik product or not. The result from prior step is the seller correctness percentage (CP) in naming their batik Solo’s product based on its pattern. The value of CP result is 82,13%. The image from marketplace’s search result also used as a data test. Then, the data test images from a specific pattern compared with image in the data set that has the same pattern to check the accuracy of image retrieval methods. The method used in this study is texture based image retrieval using SURF and Brute Force (BF). The accuracy of the SURF and BF methods is 51,32%. At last, this study rate the marketplace based on its CP and its image retrieval method accuracy value. Then it known that the marketplace with the highest CP is Lazada with 95,42% and marketplace with the best method accuracy is Bukalapak with 74,76%.
Item Type: | Thesis (Masters) |
---|---|
Additional Information: | RTMT 006.42 Lid a-1 |
Uncontrolled Keywords: | Image retrieval, SURF, motif batik Solo |
Subjects: | C Auxiliary Sciences of History > CB History of civilization |
Divisions: | Faculty of Business and Management Technology > Management Technology > 61101-(S2) Master Thesis |
Depositing User: | Berlian Rahmy Lidiawaty |
Date Deposited: | 17 Mar 2025 03:42 |
Last Modified: | 17 Mar 2025 03:42 |
URI: | http://repository.its.ac.id/id/eprint/73519 |
Actions (login required)
![]() |
View Item |