PEWARNAAN OTOMATIS CITRA GRAYSCALE BERBASIS MODEL DEEP LEARNING UNTUK PENINGKATAN INTERPRETASI VISUAL
DOI:
https://doi.org/10.29407/nr7x7m75Abstract
Pewarnaan otomatis citra grayscale menggambarkan tantangan besar dalam bidang pengolahan citra digital. Citra grayscale, meskipun mengandung informasi spasial yang melimpah, mengalami hilangnya dimensi warna yang penting, yang mungkin membatasi kemampuan interpretasi visualnya. Studi ini berorientasi pada pengembangan dan penilaian model deep learning yang ditujukan untuk mengotomatisasi proses pewarnaan gambar grayscale. Metodologi yang diterapkan mencakup pelatihan jaringan saraf tiruan konvolusional (CNN) pada dataset gambar berwarna, dengan gambar grayscale sebagai input dan gambar berwarna sebagai output yang diinginkan. Pelatihan difokuskan pada pembelajaran pemetaan yang rumit dari intensitas piksel grayscale menuju saluran warna a dan b dalam ruang warna . Hasil pengujian yang melibatkan lima gambar sampel menunjukkan bahwa model dapat menghasilkan pewarnaan yang secara visual mirip dengan citra sumber aslinya. Evaluasi kuantitatif dengan metrik Mean Absolute Error (MAE), Peak Signal-to-Noise Ratio (PSNR), dan Structural Similarity Index (SSIM) menunjukkan performa yang unggul, dengan nilai rata-rata SSIM sebesar 0.9963, menandakan kualitas pewarnaan yang Sangat Baik. Penemuan ini menekankan potensi besar deep learning dalam menghidupkan kembali informasi warna yang hilang dari gambar grayscale, membuka peluang untuk aplikasi kreatif dalam bidang restorasi foto bersejarah dan visualisasi data.
Keywords:
Deep learning, Pewarnaan Citra, PSNR, SSIM, Grayscale##plugins.themes.default.displayStats.downloads##
References
[1] R. Zhang, P. Isola, dan A. A. Efros, “Colorful image colorization,” Lect. Notes Comput. Sci. (including Subser. Lect. Notes Artif. Intell. Lect. Notes Bioinformatics), vol. 9907 LNCS, hal. 649–666, 2016, doi: 10.1007/978-3-319-46487-9_40.
[2] M. Tahmid, M. S. Alam, N. Rao, dan K. M. A. Ashrafi, “Image-to-Image Translation with Conditional Adversarial Networks,” Proc. 2023 IEEE 9th Int. Women Eng. Conf. Electr. Comput. Eng. WIECON-ECE 2023, hal. 468–472, 2023, doi: 10.1109/WIECON-ECE60392.2023.10456447.
[3] S. Iizuka, E. Simo-Serra, dan H. Ishikawa, “Let there be color!,” ACM Trans. Graph., vol. 35, no. 4, hal. 1–11, 2016, doi: 10.1145/2897824.2925974.
[4] A. Tsantekidis, N. Passalis, dan A. Tefas, “Recurrent neural networks,” Deep Learn. Robot Percept. Cogn., vol. 48, hal. 101–115, 2022, doi: 10.1016/B978-0-32-385787-1.00010-5.
[5] K. T. Li, H. R. Huang, C. C. Fan, and C. C. Lo, "Deep Image Colorization with Interactive Human Supervision," IEEE Trans. Circuits Syst. Video Technol., vol. 30, no. 2, pp. 493-506, Feb. 2020. doi: 10.1109/TCSVT.2019.2898711.
[6] Y. Lu, S. Li, L. Liu, and W. Zeng, "Deep Photo Colorization with Global and Local Attention," IEEE Trans. Image Process., vol. 29, pp. 7848-7861, 2020. doi: 10.1109/TIP.2020.3006456.
[7] Z. Ren, Z. Yan, and J. Wang, "Attention-based image colorization with multi-scale feature fusion," Neurocomputing, vol. 467, pp. 195-207, Jan. 2022. doi: 10.1016/j.neucom.2021.09.060.
[8] J. Yu, L. Ma, H. Li, Y. Liu, S. Zhang, and X. Xie, "Colorization Transformer: Global-Local Colorization with Efficient Transformer," in Proc. AAAI Conf. Artif. Intell., 2022, pp. 2951-2959. doi: 10.1609/aaai.v36i3.20191.
[9] K. Geng, H. Wu, Y. Zhang, and X. Cao, "Colorization using contrastive learning and multi-scale attention for grayscale images," Image and Vision Computing, vol. 129, p. 104603, Jan. 2023. doi: 10.1016/j.imavis.2022.104603.
[10] S. Li, W. Sun, S. Yan, W. Zhang, and L. Liu, "A Survey of Image Colorization Based on Deep Learning," IEEE Trans. Circuits Syst. Video Technol., vol. 33, no. 2, pp. 433-448, Feb. 2023. doi: 10.1109/TCSVT.2022.3197607.
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Alindro Septo Nugroho, Julian Sahertian, Rony Heri Irawan
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Copyright on any article is retained by the author(s).
- The author grants the journal, right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal’s published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
- The article and any associated published material is distributed under the Creative Commons Attribution-ShareAlike 4.0 International License
