Nondestructive Identification of Cherry-Tomato Varieties Based on Multi-Spectral Image Technology

Abstract:

Yunnan cherry-tomato and Xinjiang cherry-tomato were very similar in appearance. But they are different in taste and nutritive value. A nondestructive identification method of cherry-tomato variety based on multi-spectral image technology is proposed in this paper. Fifty Yunnan cherry-tomatoes and fifty Xinjiang cherry-tomatoes were selected, and photos were taken by Duncan MS3100 3CCD multi-spectral imager. Threshold based segmentation and mathematical morphology method were used to process the images. Nine characteristic variables were calculated to establish discriminant analysis model (DA) and least square-support vector machine model (LS-SVM). The prediction accuracy of discriminant analysis model was 72.5% and that of LS-SVM model was 80%. The results showed that LS-SVM model could identify Yunnan cherry-tomato and Xinjiang cherry-tomato well.

• Periodical: Advanced Materials Research (Volumes 108-111)
• Pages: 262-267
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.108-111.262
• Online since: May 2010
• Authors: Kai Sheng Yang, Peng Cheng Nie, Lei Feng, Yong He, Shuang Shuang Chen
• Keywords: Discriminant Analysis, Least Squares Support Vector Machines (Ls-SVM), Multi-Spectral Image, Nondestructive Identification, Threshold Segmentation

References

[1] Shao Y. N., Y. He, J. Z. Pan. Research on discrimination method of orange juice variety based on spectroscopy technology. Spectrosc. Spectr. Anal., 27, 1739-1742. (2007).

[2] He Y., X. L. Li, Y. N. Shao. Discrimination of varieties of apple using near infrared spectra based on principal component analysis and artifical neural neural network model. Spectrosc. Spectr. Anal., 26, 850-853. (2006).

[3] Liu M. H., R. F. Cheng, H. W. Lin. A study on non-destructive measurement of sugar content in navel orange using spectral imaging. Acta Agriculture Universities Jiangxiensis, 29, 443-448. (2007).

[4] Hong T. S., J. Qiao, W. Ning. Non-destructive inspection of Chinese pear quality based on hyperspectral imaging technique. Trans. CASE, 23, 151-155. (2007).

[5] Mehl P. M., Y. R. Chen, M. S. Kim. Development of hyper-spectral imaging technique for the detection of apple surface defects and contaminations. J. Food Eng., 61, 67-81. (2004).

DOI: 10.1016/s0260-8774(03)00188-2

[6] Juan X., B. Cedric, T. J. Pal. Detecting bruises on Golden Delicious, apples using hyper-spectral imaging with multiple wavebands. Biosystems Eng., 90, 27-36. (2005).

DOI: 10.1016/j.biosystemseng.2004.08.002

[7] Chen Y. R., K. Chao, M. S. Kim. Machine vision technology for agricultural applications. Comput. Electron. Agric., 36, 173-191. (2002).

[8] Wu D., D. S. Zhu, Y. He. Nondestructive detection of gray mold of eggplant based on ground multi-spectral imaging sensor. Spectrosc. Spectr. Anal., 28, 1496-1500. (2008).

[9] Yao L. J., W. M. Ding, Y. Y. Liu. Image segmentation for eggplant based on Matlab. Acta Agriculture Jiangxi, 19, 97-101. (2007).

[10] Gonzalez R. C. Digital Image Processing Using MATLAB. Publishing House of Electronics Industry, Beijing. (2005).

[11] Suykens J., G. T. Van, B. J. De. Least Squares Support Vector Machines. World Scientific, Singapore. (2002).