(3.236.118.225) 您好!臺灣時間:2021/05/16 10:20
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果

詳目顯示:::

: 
twitterline
研究生:林子超
研究生(外文):Tzu-Chao Lin
論文名稱:改良差異進化為基礎之遞迴類神經模糊網路於背光影像補償應用
論文名稱(外文):Image Backlight Compensation Using Recurrent Neuro-Fuzzy Networks Based on Modified Differential Evolution
指導教授:林正堅林正堅引用關係王德譽王德譽引用關係
指導教授(外文):Cheng-JianDe-Yu Wang
學位類別:碩士
校院名稱:朝陽科技大學
系所名稱:資訊工程系碩士班
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2008
畢業學年度:96
語文別:英文
論文頁數:49
中文關鍵詞:對比強化曝光控制模糊C平均法背光影像補償差分演算法類神經模糊網路遞迴函數鏈結類神經模糊網路
外文關鍵詞:contrast enhancementexposure controlfuzzy c-meansbacklight compensationdifferential evolutionneural fuzzy networkrecurrent functional-link-based neuro-fuzzy netw
相關次數:
  • 被引用被引用:0
  • 點閱點閱:229
  • 評分評分:
  • 下載下載:21
  • 收藏至我的研究室書目清單書目收藏:0
隨著數位相機與醫學影像數位化的普及使得數位影像成為生活中不可或缺的一部份,但在攝影時一旦曝光不足或是光源與主體的位置錯誤時,將造成背光影像的問題。在背光影像中,主體的亮度相對於周遭背景而言通常是不足的,對於影像資訊的判讀與閱覽將造成相當大的障礙而使背光影像成為無效影像,因此背光影像處理在這幾年成為相當熱門的研究課題,在許多研究中背光處理又稱為對比強化或是曝光控制。
本論文提出了一個結合模糊C平均值法、遞迴函數鏈結類神經模糊網路與改良差分演算的方法,針對不同程度的背光影像,可以做適應性的亮度對比修正,藉以獲得可供閱覽的有效影像。首先,我們將背光影像進行色彩空間轉換至YIQ,針對Y空間(亮度)的子空間進行分群與頻率統計運算求得二參數,利用已經訓練完成的遞迴函數連結類神經網路進行背光程度推論求出背光程度因子;其次,以背光程度因子來獲得適應性亮度修正曲線的係數後,即可針對背光影像各像素的亮度值進行亮度修正。
實驗的結果顯示,針對戶外及室內不同光源角度所產生的背光影像,此一方法皆能改善背光的問題。在戶外不同角度陽光所造成的背光影像中,對於樹木、雕塑、建築物、人臉等皆可使原本黑漆的主體亮度提高,且不改變周遭亮度及無邊緣效應,獲得相當不錯的效果。在戶內針對一本書籍、布偶等物品進行不同光源角度的背光影像進行處理時,背光程度皆可獲得相當不錯的改善,同時周遭環境的對比也一併改善,此一應用在未來可進一步對於夜間保全影像提供改善方法。
Popular usages of digital camera and digital medical image application have made the digital image become one part of our life. When there is the under exposure or mistaken position of light source and object, it will cause the problem of backlight image. In the backlight image, the luminance of object is low with respect to the surrounding background. It will be very difficult to read the image and to get the information from the backlight images and these images will become nonsense images. So the backlight processing has been a popular subject in these years and it is also called contrast enhancement or exposure control.
In this work, a backlight image processing combining fuzzy c-means clustering, a recurrent functional-link-based neural fuzzy network (RFLNFN), and modified differential evolution algorithm is proposed to deal with different backlight level by adaptive luminance modification for obtaining effective images. First, we transform the color space into YIQ domain, cluster the luminance of subset space Y, and calculate by the frequency statistic to get two factors. These two factors will be used for the RFLNFN model to inference the backlight degree. Then, the adaptive luminance modification function will be valid by the backlight degree for the backlight image compensation.
Our experimental results show that the contrast enhancement of backlight images which taken indoor and outdoor can be easily restored to an acceptable level with the help of the proposed model. The backlight images of tree, stabile, building, and human face taken outdoor with different angle of light sources can be enhanced well without overenhanced problem and edge effect. We also processed indoor backlight images that book or a doll for different light sources. The processed results show that the backlight of the object can be solved as well as the surrounding contrast is also promoted. This application can further be referred for the night security monitoring application.
摘 要 I
Abstract III
Preface V
Chapter 1 Introduction 1
1.1 Motivation 1
1.2 Thesis Organization 5
Chapter 2 Functional-Link-Based Neuro-Fuzzy Networks 6
2.1 Introduction 6
2.2 Structure of RFLNFN 6
Chapter 3 Supervised Evolutionary Learning for RFLNFN 11
3.1 Mutation 11
3.2 Crossover 12
3.3 Selection 13
3.4 Other variants of DE 13
3.5 Initialization Phase 15
3.6 Evaluation Phase 16
3.7 Reproduction Phase 17
3.8 Crossover Phase 17
Chapter 4 Backlight Image Processing 19
4.1 Process 19
4.2 Image Color Space Transformation 20
4.3 Factors of the Backlight Image 21
4.3 Fuzzy Inference Model 25
4.4 Image Compensation Curve 28
Chapter 5 Results 30
5.1 Input images 30
5.2 RFLNFN based on MODE 30
5.3 Backlight Image Compensation 31
Chapter 6 Conclusion 43
Reference 45
個 人 簡 歷 49


List of Figures
Fig. 1 Structure of proposed RFLNFN model. 8
Fig. 2 Process of differential evolution 11
Fig. 3 Example of 2 dimensional mutation of DE optimization 12
Fig. 4 Illustration of the crossover for 8 dimensional parameters 13
Fig. 5 Process of Differential Evolution 15
Fig. 6 Individual coding in MODE 16
Fig. 7 Contrast Enhancement Process of Backlight Images 20
Fig. 8 An example backlight image 22
Fig. 9 Frequency histogram of luminance in the backlight image 22
Fig. 10 Decumulative luminance frequency histogram of backlight image 23
Fig. 11 Architecture of basic fuzzy controller 26
Fig. 12 Fuzzy inference membership function 27
Fig. 13 Luminance compensation curve 29
Fig. 14 Square error with iterations 31
Fig. 15 Backlight image taken under a tree. 33
Fig. 16 Backlight image of an artificial iron stabile. 34
Fig. 17 Backlight image under the roof 35
Fig. 18 Backlight image of an artificial wall stabile; 36
Fig. 19 Backlight image of a book taken indoor; 38
Fig. 20 Backlight image of a book taken with different light sources; 39
Fig. 21 Backlight image of a human face 40
Fig. 22 Backlight image of Hello Kitty and camera taken indoor 41
Fig. 23 Backlight image of sunflower 42

List of Tables
Table 1 Fuzzy system control rule table 28
Table 2 Parameters set up 31
[1]M. A. Wadudx, M. H. Kabir, M. A. A. Dewan, and O. Chae, "A Dynamic Histogram Equalization for Image Contrast Enhancement," IEEE Transactions on Consumer Electronics, vol. 53, pp. 593-600, 2007.
[2]G. Zhang, P. Yan, H. Zhao, and X. Zhang, "A Contrast Enhancement Algorithm for Low-Dose CT Images Based on Local Histogram Equalization," presented at Bioinformatics and Biomedical Engineering, The 2nd International Conference on ICBBE, 2008.
[3]A. Ziaei, H. Yeganeh, K. Faez, and S. Sargolzaei, "A Novel Approach for Contrast Enhancement in Biomedical Images Based on Histogram Equalization," presented at BioMedical Engineering and Informatics, International Conference on BMEI 2008.
[4]H. Ibrahim and N. S. P. Kong, "Brightness Preserving Dynamic Histogram Equalization for Image Contrast Enhancement," IEEE Transactions on Consumer Electronics, vol. 53, pp. 1752-1758, 2007.
[5]W. Qing and R. K. Ward, "Fast Image/Video Contrast Enhancement Based on Weighted Thresholded Histogram Equalization," IEEE Transactions on Consumer Electronics, vol. 53, pp. 757-764, 2007.
[6]T. Yafei, W. Qingtao, and W. Fengjun, "Local Histogram Equalization Based on the Minimum Brightness Error," presented at Fourth International Conference on Image and Graphics, 2007.
[7]D. Menotti, L. Najman, J. Facon, and A. A. A. de Araujo, "Multi-Histogram Equalization Methods for Contrast Enhancement and Brightness Preserving," IEEE Transactions on Consumer Electronics, vol. 53, pp. 1186-1194, 2007.
[8]C.-H. Chen, C.-J. Lin, and C.-T. Lin, "A Recurrent Functional-Link-Based Neural Fuzzy System and Its Applications," presented at Proceedings of the 2007 IEEE Symposium on Computational Intelligence In Image and Signal Processing (CIISP 2007), 2007.
[9]H. M. and G. S., "Hybrid adaptive fuzzy identification and control of nonlinear systems," IEEE Transactions on Fuzzy Systems, vol. 10, pp. 198-210, 2002.
[10]L.-X. Wang, "Stable adaptive fuzzy controllers with application to inverted pendulum tracking," IEEE Transactions on Systems, Man, and Cybernetics, Part B, vol. 26, pp. 677-691, 1996.
[11]M. J. Er and S. H. Chin, "Hybrid adaptive fuzzy controllers of robot manipulators with bounds estimation," IEEE Transactions on Industrial Electronics, vol. 47, pp. 1151-1160, 2000.
[12]J. Xiu and C. Xia, "An Application of Adaptive Fuzzy Logic Controller for Switched Reluctance Motor Drive," presented at Fourth International Conference on Fuzzy Systems and Knowledge Discovery, 2007.
[13]T. R. Rangaswamy, J. Shanmugam, and K. P. Mohammed, "Model reference fuzzy adaptive controller for combustion of utility boiler," presented at SICE 2004 Annual Conference, 2004.
[14]I. del Campo, J. Echanobe, G. Bosque, and J. M. Tarela, "Efficient Hardware/Software Implementation of an Adaptive Neuro-Fuzzy System," IEEE Transactions on Fuzzy Systems, vol. 16, pp. 761-778, 2008.
[15]T. Orlowska-Kowalska and K. Szabat, "Damping of Torsional Vibrations in Two-Mass System Using Adaptive Sliding Neuro-Fuzzy Approach," IEEE Transactions on Industrial Informatics, vol. 4, pp. 47-57, 2008.
[16]W.-C. Lih, S. T. S. Bukkapatnam, P. Rao, N. Chandrasekharan, and R. Komanduri, "Adaptive Neuro-Fuzzy Inference System Modeling of MRR and WIWNU in CMP Process With Sparse Experimental Data," IEEE Transactions on Automation Science and Engineering, vol. 5, pp. 71-83, 2008.
[17]J. A. Martin, H. M. Santos, G. Farias, N. Duro, J. Sanchez, R. Dormido, S. Dormido-Canto, and J. Vega, "Dynamic Clustering and Neuro-Fuzzy Identification for the Analysis of Fusion Plasma Signals," presented at IEEE International Symposium on Intelligent Signal Processing, WISP 2007.
[18]J. Yue, J. Liu, X. Liu, and W. Tan, "Identification of Nonlinear System Based on ANFIS with Subtractive Clustering," presented at The Sixth World Congress on Intelligent Control and Automation, WCICA 2006.
[19]M. Panella and A. S. Gallo, "An input-output clustering approach to the synthesis of ANFIS networks," IEEE Transactions on Fuzzy Systems,, vol. 13, pp. 69-81, 2005.
[20]S.-N. Wu and J.-S. Wang, "An adaptive recurrent neuro-fuzzy filter for noisy speech enhancement," presented at IEEE International Joint Conference on Neural Networks, 2004.
[21]T. K. Sandra Paterlini, "Differential evolution and particle swarm optimization in partitional clustering," Computational Statics & Data Analysis, vol. 50, pp. 1220-1247, 2006.
[22]K. P. Rainer Storn, "Differential Evolution-A Simple and Efficient Heuristic for Global Optimization over Continuous Spaces," Journal of Global Optimization, vol. 11, pp. 341-359, 1997.
[23]S. Das, A. Konar, and U. K. Chakraborty, "Automatic Fuzzy Segmentation of Images with Differential Evolution," presented at IEEE Congress on Evolutionary Computation, 2006.
[24]X. Feng, A. C. Sanderson, P. P. Bonissone, and R. J. Graves, "Fuzzy Logic Controlled Multi-Objective Differential Evolution," presented at The 14th IEEE International Conference on Fuzzy Systems, 2005.
[25]H. N. Pishkenari, S. H. Mahboobi, and A. Meghdari, "On the optimum design of fuzzy logic controller for trajectory tracking using evolutionary algorithms," presented at IEEE Conference on Cybernetics and Intelligent Systems, 2004.
[26]L. Singh and S. Kumar, "Migration based parallel differential evolution learning in Asymmetric Subsethood Product Fuzzy Neural Inference System :A simulation study," presented at IEEE Congress on Evolutionary Computation, 2007.
[27]T. Takahama, S. Sakai, A. Hara, and N. Iwane, "Structural learning of neural networks by differential evolution with degeneration using mappings," presented at IEEE Congress on Evolutionary Computation, 2007.
[28]R. L. John Yen, Fuzzy Logic:intelligence, control, and information.: Prentice Hall, 1998.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top