臺灣博碩士論文加值系統

本研究的目的為研發葉材外形品質的影像檢測方法，檢測葉材的邊緣瑕疵與葉內蟲咬孔洞，檢測對象包含黃椰子葉、山蘇葉與電信蘭葉。由於此三種葉材外形特殊，呈現連續的內凹和外凸曲綫，欲以一般的外形檢測方法檢測其邊緣瑕疵有其困難度，本研究參考並修改火鶴花的外形檢測方法，使之適用於本研究上。
在邊緣檢測部份，先沿著葉緣輪廓計算以取得凹凸變化的關鍵點，再利用葉緣瑕疵處，凹凸特徵在小範圍內反複出現的特性，將之與葉材自然外形的凹凸變化作區分，即可檢測出葉緣瑕疵。
孔洞檢測方面，電信蘭葉本身具有自然孔洞，直接辨識葉內孔洞會導致誤判。由於自然孔洞與蟲咬孔洞的面積相差甚遠，在作檢測時，只要設定門檻值，排除面積大於門檻值的孔洞，即可鎖定蟲咬孔洞。

The main objective of this research is to study the image detection method of the quality of appearance of leaf material. The detection item includes the edge defect of the leaf material and pore within the leaf bitten by insects. The detection targets include yellow butterfly palm leaf, asplenium leaf and window plant leaf. Moreover, since the appearances of these three leaf materials are very special, that is, they usually show continuously concave and convex curve, hence, it would be very difficult to use general appearance detection method to detect the edge defect, however, this research has referred to and modified the appearance detection method of anthurium to make it applicable to this study.
In the edge detection aspect, calculation is made first along the profile of leaf edge to get key point of concave and convex change, then through the use of the characteristic of repeated emergence of concave and convex characteristic within small range at the defective part of the leaf edge, it is then distinguished from the concave and convex change of natural appearance of leaf material, and the defects at the leaf edge can then be detected.
In the pore detection aspect, window plant leaf itself has natural pole, hence, direct identification of the pore within the leaf will lead to wrong judgment. Since there is large area difference between natural pore and the pore bitten by insects, during the detection, only threshold value needs to be set up to exclude pores with area larger than the threshold value, and the pore bitten by insects can then be locked in.

誌謝........................................................I
摘要.......................................................II
ABSTRACT..................................................III
目錄.......................................................IV
表目錄....................................................VII
圖目錄...................................................VIII
第一章 前言.................................................1
1-1 研究動機................................................1
1-2 研究目的................................................3
第二章 文獻探討.............................................4
2-1機器視覺.................................................4
2-2 數位影像................................................4
2-2-1 色彩模型..............................................5
2-2-2 影像二值化............................................6
2-2-3 影像邊緣檢測..........................................7
2-3機器視覺在農業上之應用...................................9
第三章 材料與方法..........................................15
3-1 實驗設備...............................................15
3-2 實驗方法...............................................18
3-3 葉緣缺口判斷方法.......................................20
3-3-1 缺口判斷方法.........................................20
3-3-2 距離計算.............................................22
3-3-3 缺口判斷方法改良.....................................22
3-3-4 決定輪廓線的起始點...................................26
3-3-5 連接輪廓線...........................................30
3-3-6判斷缺口方法驗證......................................36
第四章 結果與討論..........................................46
4-1 CCD精度計算............................................46
4-2 葉緣檢測方法的參數探討.................................46
4-3 實驗結果...............................................51
4-3-1 黃椰子葉檢測結果.....................................53
4-3-2 電信蘭葉檢測結果.....................................55
4-3-3 山蘇葉檢測結果.......................................57
4-4 程式運算時間...........................................61
4-4-1 葉材檢測時間.........................................61
4-4-2 提升運算速率的探討...................................62
第五章 結論與建議..........................................68
5-1 結論...................................................68
5-2 建議...................................................71
參考文獻...................................................72
附錄A 葉材檢測數據.........................................76

1.繆紹綱。2003。數位影像處理。台灣培生教育出版股份有限公司。
2.繆紹綱。2005。數位影像處理—運用MATLAB。台灣東華書局股份有限公司。
3.王恆傑。2011。黃椰子葉外觀品質分級系統及自動化之開發。碩士論文。國立中興大學生物產業機電工程學系。
4.石坤城。2001。機器視覺應用在香菇選別系統之研究。碩士論文。國立屏東科技大學食品科學系。
5.林皇杉。2008。應用機器視覺於火鶴花切花自動分級系統之研究。碩士論文。亞洲大學資訊工程學系。
6.林福源、章加寶、林洋三、林聖泉、黃裕益。2007。機械視覺應用於花粉色澤選別機之研究。苗栗區農業改良場研究彙報1：p69-82。
7.柯建全、黃膺任、艾群。1999。應用影像處理檢測荔枝果皮之顏色變化。農業機械學刊8（1）：p59-68。
8.陳志杰。2005。應用影像處理於蓮霧外部品質檢測之應用研究。碩士論文。國立中興大學生物產業機電工程學系。
9.袁國勇、張鐵中。 2006。溫室黃瓜果實的模式識別與分割。農機化研究2006（7）：p150-153。
10.陳建文。2002。彩色影像邊緣線萃取之研究。碩士論文。國立成功大學測量工程學系。
11.黃國益、林聖泉。2000。應用機器視覺於蝴蝶蘭在大苗期 幾何特徵之估算。農業機械學刊9（2）：p13-26。
12.黃國益、林聖泉。2001。蝴蝶蘭大苗選別系統之研製。農業機械學刊10（4）：p85-98。
13.楊清富、李芳繁。1994。應用機器視覺進行番茄顏色分級之研究。農業機械學刊3（1）：p15-29。
14.鄭志祥。2002。影像次像素應用在米粒檢測之研究。碩士論文。國立中興大學農業機械工程學系。
15.维基百科。2009。邊緣檢測。網址：http://zh.wikipedia.org/wiki/邊緣檢測。上網日期：2011-07-04。
16.维基百科。2011。CIE1931色彩空間。網址：http://zh.wikipedia.org/wiki/CIE1931色彩空間。上網日期：2011-07-04。
17.维基百科。2011。HSL和HSV色彩空間。網址：http://zh.wikipedia.org/wiki/ HSL和HSV色彩空間。上網日期：2011-07-04。
18.维基百科。2011。色彩空間。網址：http://zh.wikipedia.org/wiki/色彩空間。上網日期：2011-07-04。
19.Aleixos, N., J. Blasco, E. Molto, and F. Navarron. 2000. Assessment of citrus fruit quality using a real-time machine vision system. 15th International Conference on Pattern Recognition, 482-485.
20.Hegt, H. A., R. J. D. Haye and N. A. Khan. 1998. A high performance license plate recognition system. 1998 IEEE International Conference on Systems, Man, and Cybernetics, vol. 5, 4357-4362.
21.Kamil, M.F.A., M. Mokji, U. U. Sheikh, S. A. R. Abu-Bakar. 2010. Machine Vision for Starfruit (Averrhua Carambola) Inspection. 4th Asia International Conference on Mathematical/Analytical Modelling and Computer Simulation, 333-336.
22.Mokji, M.. 2006. Concave and Convex Area on Planar Curve for Shape Defect Detection and Recognition. The 6th IEEE International Symposium on Signal Processing and Information Technology, 462-467.
23.Rahman, C. A., W. Badawy, and A. Radmanesh. 2003. A real time vehicle''s license plate recognition system. 2003 IEEE Conference on Advanced Video and Signal Based Surveillance, 163-166.
24.Seng ,W. C. and S. H. Mirisaee. 2009. A New Method for Fruits Recognition System. 2009 International Conference on Electrical Engineering and Informatics, 130-134.
25.Zeng, Q., C. Liu, Y. Miao, S. Fei, and S. Wang. 2008. A Machine Vision System for Continuous Field Measurement of Grape Fruit Diameter. 2nd International Symposium on Intelligent Information Technology Application, 1064-1068.
26.Zhang, L., Q. Yang, Y. Xun, X. Chen, Y. Ren, T. Yuan, Y. Tan, and W. Li. 2007. Recognition of greenhouse cucumber fruit using computer vision. New Zealand Journal of Agricultural Research, Vol. 50:1293-1298.
27.Zhao, J., J. Tow, and J. Katupitiya. 2005. On-tree fruit recognition using texture properties and colour data. 2005 IEEE/RSJ International Conference on Intelligent Robots and Systems, 3993-3998.
28.Zhao, Y., D. Wang, and D. Qian. 2009. Machine Vision based Image Analysis for the Estimation of Pear External Quality. The 2nd International Conference on Intelligent Systems Design and Engineering Applications, 629-632.
29.Wikipedia。2011。YIQ。網址：http://en.wikipedia.org/wiki/YIQ。上網日期：2011-07-04。