Tjahjadi, Noach (2021) Deteksi komposisi warna citra menggunakan classification supervised learning untuk mendukung machine learning pendeteksi objek = Image color composition detection using classification supervised learning to support object detection in machine learning. Bachelor thesis, Universitas Pelita Harapan.
![Title [thumbnail of Title]](http://repository.uph.edu/style/images/fileicons/text.png)
Title.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (3MB)
Preview
Abstract.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (293kB) | Preview
Preview
ToC.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (675kB) | Preview
Preview
Chapter1.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (745kB) | Preview
![Chapter2 [thumbnail of Chapter2]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter2.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (1MB)
![Chapter3 [thumbnail of Chapter3]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter3.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (1MB)
![Chapter4 [thumbnail of Chapter4]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter4.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (3MB)
![Chapter5 [thumbnail of Chapter5]](http://repository.uph.edu/style/images/fileicons/text.png)
Chapter5.pdf
Restricted to Registered users only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (224kB)
Preview
Bibliography.pdf
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (382kB) | Preview
![Appendices [thumbnail of Appendices]](http://repository.uph.edu/style/images/fileicons/text.png)
Appendices.pdf
Restricted to Repository staff only
Available under License Creative Commons Attribution Non-commercial Share Alike.
Download (3MB)
Abstract
Object detection is one of the most challenging problems in Computer Vision. It is relatively difficult because there are many variations between images that have the same object category. Other factors such as changes in perspective, scale, partial occlusion and many other things. Color is an important feature and is easiest for humans to use when viewing images. The human visual system is more sensitive to color information than "gray levels" so color was the first candidate to be used for feature extraction. However, due to the variety of appearances, lighting conditions and backgrounds, it is difficult to design a robust feature descriptor to define all types of objects.
Therefore, the writer wants to try to make a new Machine Learning Model based on the color interval value. By making this Machine Learning Model, the model will be able to predict the colors in the image. Model making begins by preparing the dataset to be used, then preprocessing the dataset, after that conduct a Classifier Training to get a reference dataset, then creating a Machine Learning Model and finally testing and evaluating will be carried out to determine the predictive performance of the Machine Learning Model created. According to the test results, the macro average of precision is 60% and the recall is 23%. These results prove that the Machine Learning Model created in this study, the Interval Model, cannot fully predict the color in the image accurately. / Object detection adalah salah satu masalah yang paling menantang dalam Computer Vision. Sulit karena banyaknya variasi antara gambar yang memiliki kategori objek yang sama. Faktor – faktor lain seperti perubahan sudut pandang, skala, oklusi parsial dan banyak hal. Warna adalah fitur yang penting dan paling mudah digunakan manusia ketika melihat gambar. Sistem penglihatan manusia lebih sensitif untuk informasi warna daripada “gray levels” sehingga warna adalah kandidat pertama yang digunakan untuk feature extraction. Namun karena keragaman penampilan, kondisi pencahayaan dan latar belakang, sulit untuk mendesain descriptor fitur yang kuat untuk menentukan semua jenis objek.
Oleh karena itu penulis ingin mencoba membuat Machine Learning Model baru berdasarkan nilai interval warna. Dengan dibuat-nya Machine Learning Model ini, model akan dapat memprediksi warna yang ada pada citra. Pembuatan Model dimulai dengan menyiapkan dataset yang akan digunakan, lalu melakukan preprocessing dataset, setelah itu melakukan Training Classifier untuk mendapatkan dataset referensi, setelah itu membuat Machine Learning Model dan akhirnya akan dilakukan testing dan evaluation untuk mengetahui performa prediksi dari Machine Learning Model yang dibuat. Menurut hasil pengujian macro average dari precision adalah 60% dan recall sebesar 23%. Hasil ini membuktikan bahwa Machine Learning Model yang dibuat dalam penelitian ini yaitu Model Interval belum bisa sepenuhnya memprediksi warna dalam citra secara akurat.
Item Type: | Thesis (Bachelor) |
---|---|
Creators: | Creators NIM Email ORCID Tjahjadi, Noach NIM01082170008 noach.nathanael@gmail.com UNSPECIFIED |
Contributors: | Contribution Contributors NIDN/NIDK Email Thesis advisor Sutrisno, Sutrisno NIDN0331126201 sutrisno.fik@uph.edu Thesis advisor Hardjono, Benny NIDN0404086401 benny.hardjono@uph.edu |
Uncontrolled Keywords: | machine learning; supervised learning; euclidean distance; color composition; color detection |
Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science |
Divisions: | University Subject > Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics Current > Faculty/School - UPH Karawaci > School of Information Science and Technology > Informatics |
Depositing User: | Users 9585 not found. |
Date Deposited: | 01 Mar 2021 09:54 |
Last Modified: | 28 Mar 2022 10:27 |
URI: | http://repository.uph.edu/id/eprint/24682 |