鴨蛋做為皮蛋及鹹鴨蛋的食品原物料，在養鴨場至食品加工廠運送途中，蛋殼可能會因此破裂而造成鴨蛋遭受汙染，在加工銷售後危害到消費者的健康。本研究在建立一套光學影像系統，檢測蛋殼是否有裂紋。
小規模加工廠檢測會以人工的方式檢測，利用敲擊來聽取聲音分辨出有無裂紋。在大規模的加工廠會採用兩種自動化技術來進行檢測，一種是使用衝擊錘敲擊傳遞波形序號的聲波檢測技術，另一種則是使用相機拍攝再對其影像進行處理的機械視覺技術。
過去機械視覺檢測技術研究中，都使用背光對鴨蛋進行照射，此方法會因蛋殼種類而產生透光率不均勻的問題。本研究在加工產品染色容許條件下對鴨蛋進行染色，以正面均勻光照射，採用機械視覺針對鴨蛋表面進行裂紋檢測，解決使用背光方式之缺點。程式開發採用 LabVIEW軟體進行撰寫，對鴨蛋影像進行處理並分類，即時標示出裂紋影像，來區分有裂紋及無裂紋影像。

Duck eggs are used as food raw materials for century egg and salted duck eggs. During the transportation from the duck farm to the food processing plant, the eggshell may crack and cause the duck eggs to be contaminated, which will endanger consumers' health after processing and sales. In this research, an optical inspection system is developed to detect cracks on the eggshells.
In small processing factories, operators use two eggs to knock each other to listen to the sound to distinguish cracks. In big processing factories, there are two automated technologies available used for crack detection. One is the acoustic wave detection technology that uses an impact hammer to transmit the waveform serial number. The other is the machine vision technology that uses a camera to grab egg images and then detect cracks from images.
In the past research on optical inspection technology, a backlight was used to irradiate duck eggs, which will cause variable light transmittance due to the type of eggshell. In this study, the duck eggs were dyed under the conditions that allowed for the dyeing of processed products, and the front surface was irradiated with uniform light. The machine vision technology was used to detect the cracks on the surface of the duck eggs, which solved the shortcomings of using the backlight method. The developed programs were written in LabVIEW software to process and classify duck eggs into good eggs or crack eggs.

[1] Oliveira C.H., Boiago M. M., Guaragni A. (2020). Effects of heat treatments and edible shell coatings on egg quality after storage at room temperature. Food Sci. Technol, Campinas, 40(1), Pages 344-348.

[2] Vollmuth T. A. (2018). Caramel color safety – An update. Food and Chemical Toxicology, Volume 111, Pages 578-596.

[3] Cantu. K, Archer G. S., Tucker. Z. S., Coufal. C. D. (2019). Effectiveness of Duck Hatching Egg Sanitization with the Combination of Hydrogen Peroxide and Ultraviolet Light. Journal of Applied Poultry Research, 28(2), Pages 301-306.

[4] 曾皇逢，2005，機械視覺教學教材之開發，碩士論文，國立屏東科技大學，機械工程系，屏東。

[5] Unay D., Gosselin B., Kleynen O., Leemans V., Destain M., Debeir. O. (2011). Automatic grading of Bi-colored apples by multispectral machine vision. Computers and Electronics in Agriculture, 75(1), Pages 204-212.

[6] Sauvola.J, Pietikäinen.M. (2000). Adaptive document image binarization. Pattern Recognition, 33(2), Pages 225-236.

[7] Kaur A., Rani U., SinghJosan G. (2020). Modified Sauvola binarization for degraded document images. Engineering Applications of Artificial Intelligence, Volume 92, Article 103672.

[8] 鄭經偉，李蒼郎，沈明輝，1998，振動頻譜應用於鴨蛋外殼裂痕檢測之研究，農業機械學刊，第7卷，第4期，43-52頁。

[9] Li Y., Dhakal S., Peng Y. (2012). A machine vision system for identification of micro-crack in egg shell. Journal of Food Engineering, 109(1), Pages 127-134.

[10] Sun K., Ma L., Pan L., Tu K. (2017). Sequenced wave signal extraction and classification algorithm for duck egg crack on-line detection. Computers and Electronics in Agriculture, Volume 142, Part A, Pages 429-439.