Please use this identifier to cite or link to this item:
標題: The Study on Estimating Volume of Rice in Granary Using Machine Vision
作者: 高嘉慶
Kao, Chia-Ching
關鍵字: Machine Vision;機器視覺;Storage,;Volume;儲穀量;體積
出版社: 生物產業機電工程學系所
引用: 1. 石坤城。2000。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。 2. 林啟昌、蔡文俊。1978。照相物理綜論。華岡出版有限公司。 3. 涂志國、徐德、譚民。2004。弧焊機器人視覺量測控制系統。機械月刊:93-98。 4. 張天生譯。2001。微積分。東華書局。 5. 許祐瑞。2004。Autodesk Inventor 8標準訓練教材。全華科技圖書股份有限公司。 6. 陳文志。2000。影像3D空間座標定位系統之研發。碩士論文。國立中央大學機械工程研究所。 7. 陳彥良。2003。即時立體視覺物體追蹤系統。碩士論文。中原大學機械工程學系。 8. 陳柏安。2003。利用電腦視覺作自走車之障礙物定位與環境掃苗。碩士論文。國立成功大學工程科學系。 9. 陳祖龍。1991。三維空間物體表面量測術。碩士論文。國立台灣大學電機工程研究所。 10. 曾靜怡。2000。XYZ三維平台中的主動式相機校正方法。碩士論文。國立中山大學機械工程研究所。 11. 黃國益。2002。應用機器視覺於蝴蝶蘭大苗幾何特徵與病害檢測之研究。博士論文。國立中興大學農業機械工程學研究所。 12. 黃敏峰。2004。人臉追蹤法應用於監控系統之研究。碩士論文。國立成功大學電機工程研究所。 13. 黃銘堂。1984。微積分與解析幾何。銀禾出版社。 14. 趙梅芳、沈邦興、吳曉明、蔣登峰。2004。多目立體視覺在工業測量中的應用研究。機械月刊:92-95。 15. 蔡玉芬。1996。應用機械視覺與類神經網路分級玫瑰切花之研究。碩士論文。國立中興大學農業機械工程學研究所。 16. 蔡奇謚。2002。利用影像扭正及地圖建構。碩士論文。國立雲林科技大學電機工程系研究所。 17. 蕭賢德。2001。電腦視覺系統應用於自走車系統障礙物檢測之研究。碩士論文。國立成功大學工程科學研究所。 18. Gonzalez, R. C. and R. E. Woods. 2002. Digital Image Processing. Prentice-Hall. 19. Huang, K. Y., T. C. Lin and J. N. Tsai. 2002. Disease Detection and Classification in Phalaenopsis Seedlings using Machine Vision. Agricultural Engineering Journal 11(1):11-30. 20. Kassim, A. A., H. Zhou and S. Ranganath. 2000. Automatic IC orientation checks. Machine Vision and Applications 12:107-112. 21. Lin, C. H. and T. Y. Lee. 2005. Metamorphosis of 3D polyhedral models using progressive connectivity transformations. Visualization and Computer Graphics, IEEE Transactions on 11(1):2 – 12. 22. Lin, I. C., J. S. Yeh and M. Ouhyoung. 2002. Extracting 3D facial animation parameters from multiview video clips. Computer Graphics and Applications, IEEE 22(6):72-80. 23. Lisa, F., J. Carrabina, C. P. Vicente, N. Avellana and E. Valderrama. 1993. Two-bit weights are enough to solve vehicle license number recognition problem. IEEE International Conference on Neural Networks 3:1242-1246. 24. Park, S. Y. and S. Murali. 2005. A multiview 3D modeling system based on stereo vision techniques. Machine Vision and Applications Journal 16(3):139-147. 25. Trucco, E. and A. Verri. 1998. Introductory Techniques for 3-D Computer Vision. Prentice-Hall. 26. Tsai, R. Y. 1987. A versatile camera calibration technique for high-accuracy 3D machine vision metrology using off-the-shelf TV cameras and lenses In IEEE Journal of Robotics and Automation, RA-3(4):323-344. 27. Zhou, Q., L. Ma, D. M. Chelberg, J. Xue, E. Peterson and M. Rowe. 2005. A novel machine vision application for analysis and visualization of confocal microscopic images. Machine Vision and Applications 16(2):96-104.
根據試驗結果,距離估算範圍為5 m-24 m,平均誤差率為2.07%。體積估算試驗分成兩部份:1. 實驗室模擬:針對五組不同體積進行估算,在水平掃描角度和縱軸掃描角度各為1°時,平均誤差率為4.61%;在水平掃描角度為3°和縱軸掃描角度為2°時,平均誤差率為9.48%。2. 實際穀倉試驗:實際穀倉儲穀量體積為235.000 m3,估算平均體積為241.367 m3,平均誤差率為2.71%。

In this study, the system for estimating volume of rice in granary using machine vision had been developed. It can estimate the volume of rice in granary as well as the volume of irregular object. The 3D surface of the object can be constructed.
The system consists of two CCD cameras, laser pointer, stepping motors and the platform. The distance between the object and the system can be estimated with 3D(three dimensional) image which are constructed from two CCD cameras' image. The rotational angles of stepping motors are used to computer the coordinates of the object. Furthermore, the volume of the object can be estimated by using the trapezoid formula. The 3D surface of the object can be plotted. The rice is piled up in granary. Therefore, the far object will not block the near object that is assumed in this study.
The simulation in laboratory and the test in granary had been undertaken. The distance between the object and the system is from 5 m to 24 m. According to the results, the averaged relative error for estimating the distance is about 2.07%. For the simulation in laboratory, five different volumes were tested. The averaged error is about 4.61% when the system scans every 1° in horizontal and vertical directions. The averaged relative error is about 9.48% when the system scans every 3° in horizontal direction and every 2° in vertical direction. For the test in granary, the estimated volume of rice is 241.367 m3 comparing to 235.000 m3 which was obtained from manual calculation. The averaged relative error is about 2.71%.
其他識別: U0005-2408200600404500
Appears in Collections:生物產業機電工程學系

Show full item record

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.