Klasifikasi Kesegaran Ikan Bandeng Berdasarkan Citra Mata Menggunakan MobileNetV3-Small Dan SVM

Kurniasari, Delia Citra (2025) Klasifikasi Kesegaran Ikan Bandeng Berdasarkan Citra Mata Menggunakan MobileNetV3-Small Dan SVM. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21081010022.-cover.pdf Download (1MB)
	Text (BAB 1) 21081010022.-bab1.pdf Download (254kB)
	Text (BAB 2) 21081010022.-bab2.pdf Restricted to Repository staff only until 19 June 2027. Download (751kB)
	Text (BAB 3) 21081010022.-bab3.pdf Restricted to Repository staff only until 19 June 2027. Download (486kB)
	Text (BAB 4) 21081010022.-bab4.pdf Restricted to Repository staff only until 19 June 2027. Download (1MB)
	Text (BAB 5) 21081010022.-bab5.pdf Download (243kB)
	Text (DAFTAR PUSTAKA) 21081010022.-daftarpustaka.pdf Download (196kB)

Abstract

Manual freshness assessment of milkfish is often subjective and inconsistent because it depends on individual experience. This study aims to develop a freshness classification model of milkfish based on fish eye images using a combination of MobileNetV3-Small as a feature extractor and Support Vector Machine (SVM) as a classification algorithm. The dataset used consists of 340 milkfish eye images with two categories, fresh and not fresh, obtained from primary data collection as well as online sources. An augmentation process was applied to expand the variety of training data, including rotation, horizontal and vertical flip, and brightness adjustment. The images were then processed through the MobileNetV3-Small architecture to produce a 576-dimensional feature vector, which became the input for the SVM model. Evaluation was conducted with various test scenarios such as variations in data proportion, kernel, and SVM parameters (C and gamma). The results showed that the combination of MobileNetV3-Small and SVM with RBF kernel, C=10, and gamma=0.01 gave the best performance with 99,02% accuracy. This model is then implemented in a web application using the Flask framework to facilitate real-time classification of fish freshness. This research shows that MobileNetV3-Small and SVM approaches are effective and efficient for visual classification of biological objects such as milkfish.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Sari, Anggraini Puspita	0716088605	anggraini.puspita.if@upnjatim.ac.id
Thesis advisor	Maulana, Hendra	1423128301	hendra.maulana.if@upnjatim.ac.id

Subjects:

T Technology > T Technology (General) > T385 Computer Graphics

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Delia Citra Kurniasari

Date Deposited:

22 Sep 2025 04:43

Last Modified:

22 Sep 2025 04:43

URI:

https://repository.upnjatim.ac.id/id/eprint/38604