Amanulhaq, Andi Aqil (2021) Implementasi Algoritma Image Hashing Dan Hamming Distance Untuk Deteksi Kemiripan Gambar. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 06111740000105-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
06111740000105-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (2MB) | Request a copy |
|
Text
06111740000105-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (2MB) | Request a copy |
Abstract
Deteksi kemiripan gambar merupakan salah satu permasalahan pada komputasi visual yang mempunyai banyak kegunaan seperti pada manajemen konten, validasi pada sistem, deduplikasi data dan sebagainya. Hal ini juga ditambah dengan penggunaan gambar digital yang sangat awam digunakan pada sistem teknologi informasi. Tugas akhir ini mengambil permasalahan deteksi kemiripan gambar pada kumpulan data gambar dalam jumlah besar. Metode algoritma image hashing dan penghitungan jarak Hamming beserta pencarian menggunakan struktur data BK-Tree digunakan untuk melakukan deteksi dan pencarian gambar mirip dengan efisien. Dari hasil tugas akhir ini, didapatkan hasil deteksi yang paling
baik pada kasus kemiripan gambar duplikat dan semi-duplikat. Distribusi jarak kedua kelas pasangan gambar tersebut berhasil terpisahkan, dimana dihasilkan skor AUC 1,00 dengan presisi sebesar 1,00 dan recall 0,99 pada nilai threshold 2. Pada kasus gambar dengan objek yang mirip, didapat nilai AUC sebesar 0,87 untuk average hash, nilai 0,76 untuk DCT Hash dan 0,74 untuk metode difference hash.
================================================================================================
Image similarity detection is one of the problems on computer vision that has many possible usages in several areas such content management, validation on a system, data deduplication and several others. The increase of demand also due to high usage of digital images on many information technology systems. This final project tries to address the problems on detecting similar images in many images data. Image hashing, Hamming distance and BK-Tree data structure are used in this final project to detect and search similar images in efficient way. From the result of this final project, we found that the methods that used to solve the problems have highest performance on detecting similar images on duplicate and semi�duplicate images with high score on 1,00 score on AUC, with precision of 1,00 and recall 0,99 with threshold value of 2. Compared to similar images due to similar objects, we found AUC scores of 0,87 for average hash, 0,76 for DCT Hash and 0,74 for difference hash.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Uncontrolled Keywords: | image hashing, jarak Hamming, deteksi, kemiripan gambar, pencarian, Hamming distance, detection, image similarity, search |
| Subjects: | Q Science > QA Mathematics Q Science > QA Mathematics > QA76.6 Computer programming. Q Science > QA Mathematics > QA9.58 Algorithms |
| Divisions: | Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis |
| Depositing User: | Andi Aqil Amanulhaq |
| Date Deposited: | 02 Sep 2021 08:29 |
| Last Modified: | 02 Sep 2021 08:29 |
| URI: | http://repository.its.ac.id/id/eprint/91267 |
Actions (login required)
 |
View Item |