Haifa Nisa Anwari, - (2025) MODIFIED SUPER-RESOLUTION GENERATIVE ADVERSARIAL NETWORK: Optimasi untuk Transmisi Gambar pada Sistem Kendaraan Listrik Otonom Berbasis LoRa. S1 thesis, Universitas Pendidikan Indonesia.
Abstract
Penelitian ini bertujuan mengoptimalkan Modified Super-Resolution Generative Adversarial Network (MSRGAN) untuk meningkatkan efisiensi transmisi gambar pada kendaraan listrik otonom berbasis teknologi Long Range (LoRa). Keterbatasan data rate pada LoRa membatasi kemampuan pengiriman gambar beresolusi tinggi, yang krusial untuk navigasi dan analisis lingkungan. MSRGAN dikembangkan untuk meningkatkan resolusi gambar berukuran 40×40 piksel menjadi 800×800 piksel tanpa kehilangan detail penting. Optimalisasi dilakukan melalui variasi jumlah dan dimensi kapsul pada generator, penerapan arsitektur residual dan non-residual, serta eksplorasi perceptual loss function. Hasil eksperimen menunjukkan bahwa modifikasi MSRGAN mampu mencapai nilai PSNR tertinggi sebesar 25,89 dB dan SSIM tertinggi sebesar 0,95, yang mencerminkan peningkatan kualitas gambar secara signifikan. Temuan ini membuktikan bahwa optimasi MSRGAN efektif dalam mendukung transmisi gambar beresolusi tinggi untuk aplikasi kendaraan listrik otonom berbasis LoRa. ----- This research aims to optimize the Modified Super-Resolution Generative Adversarial Network (MSRGAN) to enhance image transmission efficiency in Long Range (LoRa) technology-based autonomous electric vehicles. The limitation of the data rate on LoRa restricts the ability to transmit high-resolution images, which is crucial for navigation and environmental analysis. MSRGAN was developed to enhance the resolution of 40×40 pixel images to 800×800 pixels without losing important details. Optimization was carried out through variations in the number and dimensions of capsules in the generator, the application of residual and non-residual architectures, and the exploration of perceptual loss functions. The experimental results show that the modified MSRGAN is capable of achieving the highest PSNR value of 25.89 dB and the highest SSIM value of 0.95, reflecting a significant improvement in image quality. These findings prove that the optimization of MSRGAN is effective in supporting the transmission of high-resolution images for LoRa-based autonomous electric vehicle applications.
![[img]](http://repository.upi.edu/style/images/fileicons/text.png) |
Text
S_SISTEL_2104734_Title.pdf Download (690kB) |
|
Text
S_SISTEL_2104734_Chapter 1.pdf Download (59kB) |
|
Text
S_SISTEL_2104734_Chapter 2.pdf Restricted to Staf Perpustakaan Download (558kB) |
|
Text
S_SISTEL_2104734_Chapter 3.pdf Download (492kB) |
|
Text
S_SISTEL_2104734_Chapter 4.pdf Restricted to Staf Perpustakaan Download (1MB) |
|
Text
S_SISTEL_2104734_Chapter 5.pdf Download (55kB) |
|
Text
S_SISTEL_2104734_Appendix.pdf Restricted to Staf Perpustakaan Download (1MB) |
| Item Type: | Thesis (S1) |
|---|---|
| Additional Information: | ID SINTA Dosen Pembimbing Galura Muhammad Suranegara: 6703764 Ahmad Fauzi 0015098210 |
| Uncontrolled Keywords: | MSRGAN, Kendaraan listrik otonom, LoRa, Super-resolusi, Loss perseptual, Efisiensi transmisi gambar MSRGAN, Autonomous electric vehicles, LoRa, Super-resolution, Perceptual loss, Image transmission efficiency |
| Subjects: | L Education > L Education (General) T Technology > T Technology (General) |
| Divisions: | UPI Kampus Purwakarta > S1 Sistem Telekomunikasi |
| Depositing User: | Haifa Nisa Anwari |
| Date Deposited: | 05 May 2025 02:04 |
| Last Modified: | 05 May 2025 02:04 |
| URI: | http://repository.upi.edu/id/eprint/132868 |
Actions (login required)
 |
View Item |