利用高光譜影像與螢光影像偵測稻米表面酯質殘留

Observation and Measurement of Residual Bran on Milled Rice Surface Using Hyperspectral Imaging and Fluorescence Imaging

陳威同(Wei-Tung Chen)

指導教授 ： 郭彥甫

國立臺灣大學/生物資源暨農學院/生物產業機電工程學研究所/碩士(2014年)

https://doi.org/10.6342/NTU.2014.01048

摘要

稻米通常需經過精米加工程序，並且除去外層之稻殼以及米糠後才進行販售。加工之後，稻米表面之米糠殘留程度會影響稻米的品質與口感。而米糠層的主要組成成份為酯值。本研究以高光譜影像與螢光影像兩種非破壞性檢測模式檢測稻米表面之米糠殘留程度及其分佈。高光譜影像與螢光影像皆為具備空間資訊與光譜資訊之感測技術，常用於待測物上之特定化學成分定量檢測。在高光譜影像實驗中，稻米樣本經過精米以後，使用高光譜影像系統拍照，隨後將同一個稻米樣本進行染色、光學顯微鏡拍照、影像處理程序，以確知其表面實際之糠層分佈。接著使用高光譜影像作為機器學習分類器之輸入資料，建立出一個預測模型；在螢光影像實驗中，螢光指紋實驗分析結果決定了在螢光影像實驗中，所要採用的激發端與發射端之波段組合。稻米樣本使用螢光影像系統拍照，隨後進行與高光譜實驗中一樣的步驟，包含進行染色與建立機器學習之模型；最後，機器學習模型所預測之稻米表面糠層分佈將與染色法的結果進行比較。結果顯示，運用高光譜影像以及螢光光譜影像可以合理的預測出稻米上的糠層分佈。

關鍵字

稻米表面酯值殘留 ； 高光譜影像 ； 螢光影像 ； 影像處理 ； 支持向量機

並列摘要

Rice is typically consumed after milling, a process of removing the husk and bran layers on rice surface. The degree of bran residue remaining on rice surface after milling directly affects the rice quality. And the bran layer of rice mainly composed of lipids. This work proposed to nondestructively detect bran residue on single rice grain using hyperspectral imaging (HSI) and fluorescence imaging (FRI). HSI and FRI are sensing techniques that combines both spatial and spectral information and may be used for chemical compound identification and quantification. In the HSI experiment, rice samples were milled and scanned using an HSI system. Afterward, the rice samples were dyed to enable the residual bran to be identified using optical microscopy and image processing algorithms. Classifiers were then developed to predict the rice bran residue by using the HSI measurements as inputs. In the FRI experiment, appropriate combinations of fluorescence excitation and emission wavelengths were identified. Fluorescence images of rice samples at these excitation and emission wavelength combinations were then acquired and were used as the inputs to the machine learning classifier. After that, the same staining procedure and model development performed in the HSI work were applied. Bran image were predicted by using the fluorescence images. The predicted images were compared with the micrograph images for classifier performance evaluation. The proposed HSI and FRI approaches could reasonably estimate the residual bran distribution on milled rice surface.

並列關鍵字

surface lipid content ； hyperspectral imaging ； fluorescence imaging ； image processing ； support vector machine

參考文獻

Andries, E., and S. Martin. 2013. Sparse methods in spectroscopy: An introduction, overview, and perspective. Applied Spectroscopy 67(6):579-593.

Ariana, D., D. E. Guyer, and B. Shrestha. 2006. Integrating multispectral reflectance and fluorescence imaging for defect detection on apples. Computers and Electronics in Agriculture 50(2):148-161.

Arlot, S., and A. Celisse. 2010. A survey of cross-validation procedures for model selection. Statistics Surveys 4:40-79.

Bhattacharya, D. R., and C. M. Sowbhagya. 1972. A colorimetric bran pigment method for determining the degree of milling rice. . Sci. Food Agriculture 23:161-169.

Chang, C.-C., and C.-J. Lin. 2011. LIBSVM: A library for support vector machines. ACM Trans. Intell. Syst. Technol. 2(3):1-27.