請用此 Handle URI 來引用此文件: http://hdl.handle.net/11455/93624
標題: Application of Machine Vision of Hatching Eggs Images of Research
應用機器視覺探討種蛋孵化影像之研究
作者: 曾國豪
陳鴻毅
鄭經偉
Guo-Hao Zeng
Hung-Yi Chen
Ching-Wei Cheng
關鍵字: 
孵化
機器視覺影像
Eggs
Hatching
Machine vision
出版社: 臺中巿: 國立中興大學農學院
摘要: The purpose in this study was investigative of the feasibility by digital image used in egg hatching detection, and compared the difference in image capture view for the detection of egg hatching. Chose the Lohmann eggs as experimental subjects, and captured the profile and bottom view of the egg images daily by image capture system, that images were analyzed by RGB model. In the result show that the R, G value of the normal hatching eggs image decreased with the number of days increased. B value have unrelated with hatching. Established discriminant index with hatching eggs R, G value in the experimental group. Experimental results show that the later period suspension eggs could not be determined, and G value were better than R value, the bottom images were better than profile images. At 6th day, the bottom images used G value to detection hatching eggs and early period suspension eggs that the accuracy were 94% and 100%. The method of this study can be effectively applied to the monitoring and early detection of hatching eggs.
本研究主要的目為探討數位影像應用於種蛋孵化程度檢測之可行性,並分析取像角度對於種蛋孵化檢測之差異。以龍門蛋雞種蛋作為實驗對象,每日以影像擷取系統分別擷取種蛋側面及底面之照蛋影像,並以影像RGB模型進行分析。實驗結果顯示,正常孵化種蛋影像之R、G值之變化隨孵化天數增加而降低,而B值並不具判別意義。並以實驗組之出雛蛋R、G值建立判別指標,其結果顯示後期中止蛋皆無法有效判別,而出雛蛋及初期中止蛋G值之判別準確率較R值為佳,而底面取像較側面取像有較好之準確率。以第六天底部取像並以G值作為判別,其出雛蛋及前期中止蛋之判別準確率可達94%及100%,本研究之結果可有效應用於種蛋孵化之監測及孵化初期之選別作業。
URI: http://hdl.handle.net/11455/93624
顯示於類別:第62卷 第01期

文件中的檔案:
檔案 大小格式 
85456-6.pdf1.02 MBAdobe PDF檢視/開啟


在 DSpace 系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。