臺灣博碩士論文加值系統

本篇文章主要探討以直方圖為基礎的技術，提高影像對比度的方法，本篇文章針對兩種需求不同的影像，主體特色影像、景色影像。發展出對應的對比強化效果良好處理法，並透過實驗數據證實效果良好。並展望此提高對比影像演算法可以應用於即時影像系統、智慧手機攝影系統、數位相機系統。
第一種方法是【可適性雙段直方圖等化法】，對於主體-背景區分鮮明的特寫影像，以強調主體對比提高為目標。發展出一種基於Bi-HE下新的影像處理方法，可以將影像主體與背景成分區分開來，擷取影像的灰階成份分析，提高主體影像灰階範圍值，縮小背景影像灰階範圍但是保留背景影像辨識度，充分提高主體影像對比度，發揮特寫影像的凸顯效果。透過參數shift實驗與調整，明確掌握主體影像對比提高幅度。
第二種方法是【局部統計資訊直方圖等化法】，對於一般景色影像，以強調整體對比度提高，發展出以區域資訊為基礎，從提高每個區域的對比值，進而將整體影像對比度提高的方法，透過width、diff參數調整計算的windows大小，歸納分析影像類型，以平均區域變異數為指標，對各種不同類型景色影像得到良好對比提高的效果。

Many image enhancement methods have been proposed. Histogram equalization is a very popular approach. Some histogram-based methods have been proposed to improve the performance of enhancement under different conditions. In this thesis, we propose two new histogram-based image enhancement methods with different criteria.
Our first method is the Adaptive Bi-Segment Histogram Equalization (ABHE). In this method, the foreground pixels and background pixels are separated by image segmentation based on the histogram. Then, we assign a smaller range of gray levels to background pixels and a larger range to the foreground pixels. Finally, the histogram equalization is applied to each range separately. Experimental results show that the objects in the foreground are enhanced, so that we can recognize more details of them.
Our second method is the Local statistic information Histogram Equalization (LSIHE). In this method, we focus on the following two factors. Firstly, the image enhancement method should be robust to noise. Due to noise, the image may have +5 to -5 additive gray level variation. We treat the low level variation as the noise and it will not be expended in the histogram equalization process. Secondly, expending the contrast between distant pixels is not as important as the neighboring pixels. For human eyes, the contrast between distant pixels has no significant effect on image enhancement. Based on the two criteria, we propose a new algorithm to enhance the images. Experimental results show that our method is satisfactory.

摘要 i
ABSTRACT ii
誌謝 iii
目錄 iv
圖目錄 vi
表目錄 ix
第一章 緒論 1
1.1 研究動機 1
1.2 研究目的 2
1.3 研究方法 3
第二章 文獻回顧 5
2.1 灰階影像矩陣 5
2.2灰階直方圖 5
2.3 直方圖等化 6
2.4 Bi-HE 12
2.5 Multi-HE 12
2.6 Local HE 13
2.7 Matlab影像處理 13
2.8 Matlab中的彩色影像矩陣 14
2.9邊緣偵測與sobel operator 17
2.10 otsu's method 20
第三章 可適性雙段直方圖等化法 22
3.1從彩色影像轉換為YCbCr影像 22
3.2 otsu'method為基礎 直方圖切割 23
3.3 基於BI-HE主體影像切割與等化 23
3.4 灰階映射區域調整 28
3.5可適性雙段直方圖等化法 30
3.6使用不同shift產生的結果 33
3.7 評量指標 35
3.8研究結果 37
3.9 選取最適當的shift值達到最佳效果 48
3.10 ABHE結論 54
第四章 局部統計資訊直方圖等化法 55
4.1 將RGB色域轉換為YCbCr色域 56
4.2 影像處理window結構 57
4.3 透過映射曲線得到結果影像 61
4.4 平均區域變異數指標 67
4.5 研究結果 68
4.6實驗結果的平均區域變異數變化趨勢 93
4.7各類型影像實驗結果趨勢分析 100
4.8 LSIHE結論 102
第五章 結論 104
5.1 綜合結論 104
5.2 未來展望 105
參考文獻 106

[1] Sonka, M. , Hlavac, V. , & Boyle, R. (2008). Image Processing Analysis and Machine Vision(3rd ed. ). Toronto :Thomson Learning.
[2] Zhang, H. , Li, Y. , Chen, H. , Yuan, D. , & sum, M. (2013). Perceptual contrast enhancement withdynamic range adjustment. Optik, 124, 5906-5913.
[3] Khan, M. F. , Khan, E. , & Abbasi, Z. A. (2014). Segment selective dynamic histogram equal-ization for brightness preserving contrast enhancement of images. Optik, 125, 1385-1389.
[4] Gonzalez, R. C. , & Woods, R. E. (2009). Digital Image Processing (3ed). Prentice Hall.
[5] kim, Y. T. (1997). Contrast enhancement using brightness preserving bi-histogram equalization. IEEE Trans Consum Electron , 43, 1-8.
[6] Wan, Y. , Chen, Q. , & Zhang, B. M. (1999). Image enhancement based on equal area dualistic sub-image histogram equalization method. IEEE Trans Consum. Electron, 45, 68-74.
[7] Chen, S. D. , & Ramli, A. R. (2003). Minimum mean brightneww error bi-histogram equalization in contrast enhancement. IEEE Trans. consum . Electron , 49, 1310-1319
[8] Chen, S. D. , & Ramli, A. R. (2003). Contrast enhancement using recursive mean-separate histogram equalization for scalable brightness preservation. IEEE Trans. Con-sum. Electron, 49, 1301-1309.
[9] Sim, K. S. , Tso, C. P. , & Tan, Y. Y. (2007). Recursive sub-image histogram equalization applied to gray scale images. Pattern Recognit Lett , 28, 1209-1221.
[10] Kim, Y. T. (1997). Contrast enhancement using brightness preserving bi-histogramequalization. IEEE Trans. Consum. Electron, 43, 1-8.
[11] Wan, Y. , Chen, Q. , & Zhang, B. M. (1999). Image enhancement based on equal area dualisticsub-image histogram equalization method. IEEE Trans. Consum. Electron, 45, 68-75
[12] Sim, K. S. , Tso, C. P. , & Tan, Y. Y. (2007). Recursive sub-image histogram equalization appliedto gray scale images. Pattern Recogn Lett, 28, 1209-1221.
[13] Ooi, C. H. , Kong, N. S. P. , & Ibrahim, H. (2009). Bi-histogram equalization with a plateaulimit for digital image enhancement. IEEE Trans Consum Electron, 55, 2072-2080
[14] Dale-Jones, R. , & Tjahjadi, T. (1993). A study and modification of the local histogram equalization algorithm. PatternRecognition26, 9, 1373-1381.
[15] Kim, T. K. , Paik, J. K. , & Kang, B. S. (1998). Kang, Contrast enhancement system using spatially adaptive histogram equalization with temporal filtering. IEEE Transactionson Consumer Electronics44, 1, 82-87.
[16]Pizer, S. M. , Amburn, E. P. Austin, J. D. , Cromatie, R. , Gesslowitz, A. , Greer, T. , Romeny, B. H. , Zimmerman, J. B, & Zuiderveld, K. (1987). Adaptive histogram equalization and its variations. Comput Vis Graph Image Process, 39, 355-368.
[17] Pizer, S. M. , Johnston, R. E. , Ericksen, J. P. , Yankaskas, B. C. , & Muller, K. E. (May 1990). Contrast-limited adaptive histogram equalization: speed and effectiveness. Vis Biomed Comput, 337-345.
[18] Han, J. H. , Yang, S. J. , & Lee, B. U. (2011). A novel 3-D color histogram equalization methodwith uniform 1-D gray scale histogram. IEEE Trans Image Process, 22, 506-512.
[19] Mlsna, P. A. , Zhang, Q. , & Rodrguez, J. J. (1996). 3-D histogram modification of color images. Proc of Int Conf on Image Process, 3, 1015-1018.
[20] Zhang, Q. , Mlsna, P. A. , & Rodriguez, J. J. (1996). A recursive technique for 3-D histogramenhancement of color images. Proc of IEEE Southwest Symposium on ImageAnalysis and Interpretation, 218-223.
[21] Otsu, N. (1979). A threshold selection method from gray-level histograms. IEEE Trans System Man Cybern, 9, 62–66.
[22]陳逸勤、洪集輝(2013)。Local statistic Information based Histogram Equalization。第十二屆離島資訊技術與應用研討會，117。
[23] 陳逸勤、洪集輝(2014)。範圍調整雙區域直方圖等化。第十三屆離島資訊技術與應用研討會，62。