Laili, Ahmad Athfi Noor (2020) Analisis Performa Sistem Deteksi Objek Konstruksi Bangunan Menggunakan Convolutional Neural Network. Other thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
07111540007004-Undergraduate_Thesis.pdf Download (2MB) | Preview |
Abstract
Deep learning merupakan salah satu cabang dari machine learning yang memungkinkan komputer dapat belajar dari data dan pengalaman. Salah satu algoritma dalam deep learning adalah Convolutional Neural Network (CNN) dimana sering digunakan dalam deteksi objek. Namun, terdapat beberapa kendala antara lain jika objek pada gambar yang dideteksi terhalang objek lainnya, akurasi yang dihasilkan mengalami penurunan dan resolusi yang rendah dapat mengakibatkan objek tidak terdeteksi. Oleh karna itu, dalam penelitian ini digunakan metode Faster R-CNN untuk meningkatkan performa sistem deteksi objek dengan menggunakan dua model CNN yakni Resnet50 dan VGG-16. Hasil dari pengujian yang diperoleh dengan menggunakan model ResNet50 mempunyai kesalahan 0,3471 dengan tingkat akurasi sebesar 85,33%. Sedangkan hasil pengujian dengan model VGG-16 memiliki kesalahan sebesar 0,6782 dengan akurasi sebesar 93,39%.
=================================================================================================================================
Deep learning is one of machine learning that enables computers to learn from data and experience. One of the algorithm in machine learning is Convolutional Neural Network (CNN) which is often used in object detection. However, there are several problem include when the object is blocked to another object, the result of accuracy is decreases and low resolution make the object not being detected. Therefore, in this research the R-CNN Faster method was used to measure the performance of object detection systems using two CNN models namely Resnet50 and VGG-16. The results of tests obtained using the ResNet50 model have an error of 0,3471 with an accuracy rate of 85,33%. While the test results with the VGG-16 model have an error of 0,6782 with an accuracy of 93,39%
| Item Type: | Thesis (Other) |
|---|---|
| Additional Information: | RSE 006.42 Lai a-1 2020 |
| Uncontrolled Keywords: | Convolutional neural network, klasifikasi gambar, ResNet50, VGG-16. |
| Subjects: | T Technology > TA Engineering (General). Civil engineering (General) > TA1637 Image processing--Digital techniques |
| Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Electrical Engineering > 20201-(S1) Undergraduate Thesis |
| Depositing User: | Ahmad Athfi Noor Laili |
| Date Deposited: | 06 Jun 2024 07:36 |
| Last Modified: | 06 Jun 2024 07:36 |
| URI: | http://repository.its.ac.id/id/eprint/73927 |
Actions (login required)
 |
View Item |