在最新的影像壓縮技術MPEG-4及MPEG-7中，從影像中辨識出主要的VOs是相當重要的一環。在本篇論文中，我們成功的將Neuro-Fuzzy modeling技術應用於影像臉部及身體的擷取。我們首先利用Fuzzy Clustering的技術將影像中的pixels分類成數個clusters，並擷取出Fuzzy rules。接著利用人類臉部的膚色及身體的移動量等特性，判斷出clusters屬於前景或背景。最後，利用Neural network修正Fuzzy rules，將前景及背景間的邊緣地帶做細部的判斷，完成整張影像人體的擷取。由於我們的方法能更精確的取出Neural network所需要的training data，因此能大幅提高臉部及身體擷取的正確性。此外，在VOs移動量不大的情況下，我們的方法仍然可以成功的擷取出所要的VOs。

For some new multimedia applications using Mpeg-4 or Mpeg-7 video coding standards, it is important to find the main objects in a video frame. In this thesis, we propose a neuro-fuzzy modeling approach to the detection of human face and body. Firstly, a fuzzy clustering technique is performed to segment a video frame into clusters to generating several fuzzy rules. Secondly, chrominance and motion features are used to roughly classify the clusters into foreground and background, respectively. Finally, the fuzzy rules are refined by a fuzzy neural network, and the ambiguous regions between foreground and background are further distinguished by the fuzzy neural network. Our method improves the correctness of human face and body detection by getting training data more precisely. Besides, we can extract the VOs correctly even the VOs have no obvious motion in the video sequence.

目錄

摘要……………………………………………………...i
Abstract………………………………………………....ii

第一章 簡介…………...…………………...………..…1

1.1 Video Objects....................................…………………...…….…........2
1.2 Object-based Compression.............................….………………….....4

第二章 Objects Segmentation...…...........…....…...........6

2.1 Introduction To Objects Segmentation.....……………..…...............6
2.2 Segmentation Of Human Face......................……………..…...........7
2.3 Segmentation Of Human Body.........................………..……….......8
2.4 Introduction To Nikolaos Doulamis' Method....………..……..........9
2.5 Introduction To Kompatsiaris’Method………………………........14
2.6 Drawbacks Of these Two Methods...........………………...............17

第三章 Fuzzy Neural Network之簡介..……………...19

3.1 Architecture of the neuro-fuzzy network..............…………….….19
3.2 Hybrid Learning Algorithm......................…………………...........22

第四章 Neuro-Fuzzy Model 應用於 Human Face and Body Detection.…………………………………...…...29

4.1 Self-Contructing Clustering……..............……………….....…......29
4.2 重新編號及合併過小的區塊...........…………...…..........….........32
4.3 Human Face Extraction.............................……………...…….......34
4.4 Human body Extraction......................…………………….............39
4.5 Refining Fuzzy Rules by Neural Fuzzy Network……...........….....43

第五章 模擬結果…………………………………..…45

5.1 實際影像模擬結果..……………………........................….............46
5.2 數據比較分析.........................................…………………………..51

第六章 總結...........................................………….......52

參考文獻

[1] Doulamis, N. D. and Doulamis, A. D. and Kollias, S. D., “Object Based Coding of Video Sequences at Low Bit Rates Using Adaptively Trained Neural Networks,” in Proceeding of the 6th IEEE International Conference on Electronics, Circuits and Systems, pp. 969 –972, 1999.

[2] Kim, S. H. and Kim, N. K. and Ahn, S. C. and Kim, H. G., “Object Oriented Face Detection Using Range and Color Information,” in Proceedings of the 3th IEEE International Conference on Automatic Face and Gesture Recognition, pp. 76-81, 1998.

[3] Kim, S. H. and Kim, H.G., “Face Detection Using Multi-modal Information,” in Proceeding of the 4th IEEE International Conference on Automatic Face and Gesture Recognition, pp. 14-19, 2000.

[4] Kuo, C. M. and Hsieh, C. H. and Huang, Y. R., “A New Temporal-Spatial Image Sequence Segmentation for Object-Orented Video Coding,” in Proceeding of IEEE Asia Pacific Conference on Multimedia Technology and Applications, pp. 117-127, 2000.

[5] Wang, H. and Chang, S.-F., “A Highly Efficient System for Automatic Face Region Detection in MPEG Video,” in IEEE Transactions on Circuits and Systems for Video Technology, Vol 7, NO. 4, AUGUST 1997.

[6] Doulamis, N. and Doulamis, A. and Kollias, S., “Improving the Performance of MPEG Compatible Encoding at Low Bit Rates Using Adaptive Neural Networks,” in Real-Time Imaging, Vol 6, NO. 5, 2000.

[7] Kompatsiaris, I. and Strintzis, M. G., “Spatiotemporal Segmentation and Tracking of Objects for Visualization of Videoconference Image Sequence,” in IEEE Transactions on Circuits and Systems for Video Technology, Vol. 10, NO. 8, DECEMBER 2000.

[8] Ouyang, C.-S. and Lee, S.-J., “A New Self-Constructing Approach for Neuro-Fuzzy Modeling,” in Proceedings of the 8th International Fuzzy Systems Association World Congress, pp. 787-791, Taipei, Taiwan, 1999.

[9] Ouyang, C.-S. and Lee, S.-J., “An Improved Learning Algorithm for Rule Refinement in Neuro-Fuzzy Modeling,” in Proceedings of the 3rd International Conference on Knowledge-Based Intelligent Information Engineering Systems, pp. 238-241, Adelaide City, Australia, 1999.

[10] Chai, D. and Ngan, K. N., “Face Segmentation Using Skin-Color Map in Videophone Applications,” in IEEE Transactions on Circuits and Systems for Video Technology, Vol. 9, NO.4, JUNE 1999.