請用此 Handle URI 來引用此文件：
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92326完整後設資料紀錄
| DC 欄位 | 值 | 語言 |
|---|---|---|
| dc.contributor.advisor | 傅楸善 | zh_TW |
| dc.contributor.advisor | Chiou-Shann Fuh | en |
| dc.contributor.author | 方郁婷 | zh_TW |
| dc.contributor.author | Yu-Ting Fang | en |
| dc.date.accessioned | 2024-03-21T16:37:57Z | - |
| dc.date.available | 2024-03-22 | - |
| dc.date.copyright | 2024-03-21 | - |
| dc.date.issued | 2024 | - |
| dc.date.submitted | 2024-01-30 | - |
| dc.identifier.citation | [1] L. V. Hevia, M. A. Patricio, J. M. Molina, and A. Berlanga, “Optimization of the ISP Parameters of a Camera Through Differential Evolution,” IEEE Access, Vol. 8, pp. 143479-143493, doi: 10.1109/ACCESS.2020.3014558, 2020.
[2] Image Engineering, “Multipurpose Test Chart for High-Speed Camera Testing,” https://www.image-engineering.de/products/charts/all/425-te42, 2023. [3] ISO, “ISO 12233:2000(en) Photography — Electronic Still-Picture Cameras — Resolution Measurements,” https://www.iso.org/obp/ui/#iso:std:iso:12233:ed-1:v1:en, 2000. [4] ISO, “ISO 12233:2017(en) Photography — Electronic Still Picture Imaging — Resolution And Spatial Frequency Responses,” https://www.iso.org/obp/ui/#iso:std:iso:12233:ed-3:v1:en, 2017. [5] J. Jain, Y. Zhou, N. Yu, and H. Shi, “Keys to Better Image Inpainting: Structure and Texture Go Hand in Hand,” arXiv:2208.03382v2, 2022. [6] D. Karaboga and B. Basturk, “Artificial Bee Colony (ABC) Optimization Algorithm for Solving Constrained Optimization Problems,” Proceedings of the International Fuzzy Systems Association World Congress, Cancun, Mexico, pp. 789–798, 2007. [7] Y. Kim, J. Lee, S. S. Kim, C. Yang, T. H Kim, and J. S. Yim, “DNN-Based ISP Parameter Inference Algorithm for Automatic Image Quality Optimization,” Proceedings of Symposium on Electronic Imaging, Society for Imaging Science and Technology, https://doi.org/10.2352/ISSN.2470-1173.2020.9.IQSP-315, 2020. [8] J. C. Lagarias, J. A. Reeds, M. H. Wright, and P. E. Wright, “Convergence Properties of the Nelder-Mead Simplex Algorithm in Low Dimensions,” SIAM Journal on Optimization, Vol. 9, No. 1, pp. 112-147, 1998. [9] J. Nishimura, T. Gerasimow, S. Rao, A. Sutic, C. T. Wu, and G. Michael, “Automatic ISP Image Quality Tuning Using Nonlinear Optimization,” https://arxiv.org/abs/1902.09023, 2018. [10] R. Olaf, F. Philipp, and B. Thomas, “U-Net: Convolutional Networks for Biomedical Image Segmentation,” Lecture Notes in Computer Science, Vol. 9351. pp. 234-241, 2015. [11] H. S. Park, “Architectural Analysis of a Baseline ISP Pipeline,” In: Kyung, CM. (eds) Theory and Applications of Smart Cameras, KAIST Research Series, Springer, Dordrecht, https://doi.org/10.1007/978-94-017-9987-4_2, 2016. [12] K. Tero, L. Samuli, A. Miika, H. Janne, L. Jaakko, and A. Timo, “Analyzing and Improving the Image Quality of StyleGAN,” Proceedings of Computer Vision and Pattern Recognition, Virtual, pp. 8107-8116, doi: 10.1109/CVPR42600.2020.00813, 2020. [13] E. Tseng, F. Yu, Y. T. Yang, F. Mannan, A. S. Arnaud, D. Nowrouzezahrai, J. F. Lalonde, and F. Heide, “Hyperparameter Optimization in Black-Box Image Processing Using Differentiable Proxies,” ACM Transactions on Graphics, Vol. 1, No. 1, pp. 1-14, 2019. [14] R. Zhang, P. Isola, A. A. Efros, E. Shechtman, and O. Wang, “The Unreasonable Effectiveness of Deep Features as a Perceptual Metric,” Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, UT, pp. 586-595, doi: 10.1109/CVPR.2018.00068, 2018. | - |
| dc.identifier.uri | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/92326 | - |
| dc.description.abstract | ISP (Image Signal Processor)中文為影像訊號處理器，當光進入手機鏡頭被感應器轉成訊號後，需要經由一系列的影像改進算法(例如白平衡、去噪和去馬賽克以及其他影像增強算法)來將訊號轉變成滿足特定應用或需求的影像，由於影像轉換的複雜性，ISP通常具有許多參數需要調整。
ISP 調參是智慧型手機設計與開發中的一個重要階段，因為手機的硬體特性不同，接收到的光訊號也不同，因此在出廠前都需要調整ISP內的參數，目的是為了讓影像達到最高的品質，這些參數通常由具有經驗的專家手動調整，需要不停地反覆測試來找到最佳的參數，因此調整時間可達數周甚至數個月。 本篇論文提出的方法是以StyleGAN2模型為基礎，結合U-Net的架構，藉由深度學習來達到自動化推薦ISP參數，減少ISP調參所需的人力，並且將調整參數的時間減少到幾個小時以內。 | zh_TW |
| dc.description.abstract | ISP (Image Signal Processor) is a processor that processes the signal from light into images. When light enters the mobile phone camera and is converted into a signal by the sensor, it needs to undergo a series of image algorithms (such as white balance, noise reduction, mosaic removal, and other image enhancement algorithms) to transform the signal into an image that meets specific applications or requirements. Due to the complexity of image transformation, ISP typically has many parameters that require adjustment.
ISP tuning is a crucial stage in the design and development of smartphones. Due to variations in hardware characteristics among smartphones, the received optical signals differ. Therefore, adjustments to the parameters within the ISP (Image Signal Processor) are necessary prior to manufacturing, aiming to achieve the highest image quality possible. Typically, these parameters are manually adjusted by experienced experts, involving a process of continuous testing to find the optimal settings. Consequently, the tuning process can take several weeks or even months. Our method is based on the StyleGAN2 (style Generative Adversarial Network 2) model and incorporates the U-Net architecture. Through deep learning, we aim to automate the recommendation of ISP parameters, reducing the human effort required for tuning. This approach also significantly decreases the time needed to adjust parameters to within a few hours. | en |
| dc.description.provenance | Submitted by admin ntu (admin@lib.ntu.edu.tw) on 2024-03-21T16:37:57Z No. of bitstreams: 0 | en |
| dc.description.provenance | Made available in DSpace on 2024-03-21T16:37:57Z (GMT). No. of bitstreams: 0 | en |
| dc.description.tableofcontents | 誌謝 i
中文摘要 iii ABSTRACT iv CONTENTS vi LIST OF FIGURES ix LIST OF TABLES xiii Chapter 1 Introduction 1 1.1 Overview 1 1.2 ISP 1 1.3 ISP Parameters 5 1.4 ISP Parameters Rules 8 1.5 Image Quality 12 1.5.1 Standard Charts 12 1.5.2 Real Scene 15 1.6 ISP Tuning 16 1.7 Thesis Organization 17 Chapter 2 Related Works 18 2.1 Overview 18 2.2 Optimization Algorithms 18 2.3 Machine Learning Methods 22 Chapter 3 Methodology 26 3.1 Overview 26 3.2 Recommendation Pipeline 27 3.3 Dataset 28 3.3.1 Training Dataset 28 3.3.2 Recommended Dataset 29 3.4 Dataset Alignment 29 3.5 Select ROI (Region Of Interest) 30 3.6 Train Model 31 3.6.1 Stage 1: Train an ISP Proxy Model 32 3.6.2 Stage 2: Optimizing the ISP Parameters 33 3.7 Loss Functions 34 Chapter 4 Experiment Results 37 4.1 Overview 37 4.2 Scene 1 38 4.2.1 PD (Perceptual Distance↓) 39 4.2.2 Voting 43 4.3 Scene 2 46 4.3.1 PD (Perceptual Distance) 47 4.3.2 Voting 51 4.4 Scene 3 54 4.4.1 PD (Perceptual Distance) 55 4.4.2 Voting 59 4.5 Scene 4 63 4.5.1 PD (Perceptual Distance) 64 4.5.2 Voting 68 4.6 Time Consumption 71 Chapter 5 Conclusion and Future Works 73 References 75 | - |
| dc.language.iso | zh_TW | - |
| dc.subject | 參數推薦 | zh_TW |
| dc.subject | ISP | zh_TW |
| dc.subject | StyleGAN2 | zh_TW |
| dc.subject | U-Net | zh_TW |
| dc.subject | 深度學習 | zh_TW |
| dc.subject | Deep Learning | en |
| dc.subject | U-Net | en |
| dc.subject | StyleGAN2 | en |
| dc.subject | Parameter Recommendation | en |
| dc.subject | ISP | en |
| dc.title | 方參數:影像訊號處理器參數推薦基於風格生成對抗網路與U網路 | zh_TW |
| dc.title | FangParameter: Image Signal Processor Parameter Recommendation Based on StyleGAN2 and UNet | en |
| dc.type | Thesis | - |
| dc.date.schoolyear | 112-1 | - |
| dc.description.degree | 碩士 | - |
| dc.contributor.oralexamcommittee | 李政傑;方瓊瑤 | zh_TW |
| dc.contributor.oralexamcommittee | ZHENG-JIE LI;QIONG-YAO FANG | en |
| dc.subject.keyword | ISP,參數推薦,深度學習,U-Net,StyleGAN2, | zh_TW |
| dc.subject.keyword | ISP,Parameter Recommendation,Deep Learning,U-Net,StyleGAN2, | en |
| dc.relation.page | 77 | - |
| dc.identifier.doi | 10.6342/NTU202400230 | - |
| dc.rights.note | 未授權 | - |
| dc.date.accepted | 2024-02-01 | - |
| dc.contributor.author-college | 電機資訊學院 | - |
| dc.contributor.author-dept | 資訊工程學系 | - |
| 顯示於系所單位： | 資訊工程學系 | |
文件中的檔案：
| 檔案 | 大小 | 格式 | |
|---|---|---|---|
| ntu-112-1.pdf 未授權公開取用 | 6.11 MB | Adobe PDF |
系統中的文件，除了特別指名其著作權條款之外，均受到著作權保護，並且保留所有的權利。