Design Investigation Quality System of Rice Grain Phenotype by Using Digital Images Combined with Machine Learning

Main Article Content

Kulwadee Tanwong Poonpong Suksawang Yunyong Punsawad Chalatip Thumkanon

Abstract

        This research was aimed to design inspection system of the rice grain phenotypic quality, by using digital images combining machine learning. The research methodology divided into two part were hardware, and software. The research objective was developed rice grain phenotypic quality inspection system prototype. Hardware development was created shooting device for reducing noise, and burden for improving image quality before importing images into the process.  And created geometric reference ruler to define the object size on the image, set to width of 0.5 centimeters, length of 1.5 centimeters. Software development was to created system to check the rice grain phenotype quality by using digital images combined with machines learning for classifying rice grain quality group according to the Thai jasmine rice standard product criteria under the Ministry of Commerce 2016. The research found that, the classification between the full rice grain and the stomach egg rice grain with only a small amount of egg contents was only characteristic that makes the discriminant characteristics of both naked eye specialist and the system designed.

Keywords

Article Details

Section
Research Article

References

[1] National Bureau of Agricultural Commodity and Food Standards (ACFS), “Thai Jasmine Rice,” (In Thai). ACFS 4000-2017.
[2] The rice trader, “10th The Rice Trader World Rice Conference.” [Online]. Available: https://thericetrader.com/conferences/2018-wrc-hanoi/worlds-best-rice/. [Accessed: 14-Nov-2018].
[3] Thai PBS News, "Thai Jasmine Rice 105 get the best taste of rice in the world.” (In Thai). [Online]. Available: https://news.thaipbs.or.th/content/267574. [Accessed: 9-Nov-2018].
[4] THAIBIOTECH.INFO. “What is Phenotype?.” (In Thai). [Online]. Available: https://www.thaibiotech.info/what-is-phenotype.php. [Accessed: 5-Feb-2018].
[5] Committee drafts strategic Foreign Agriculture Ministry of Agriculture and Cooperatives, Strategic Foreign Agriculture Ministry of Agriculture and Cooperatives 2017-2021. (In Thai). Bangkok: Office of Agricultural Economics, 2017.
[6] Export Commodity Standards Act, Subject: Standard product Thai Jasmine Rice and standard for Thai Jasmine Rice (Issue 3) 2016 Page 5, book 133, special episode 243 d, Bangkok: Ministry of Commerce, 2016.
[7] V. D. Daygon et al., “Understanding the Jasmine phenotype of rice through metabolite profiling and sensory evaluation,” Metabolomics, vol. 12, no. 4, pp. 63, Mar. 2016.
[8] S. Tilley and H. J. Rosenblatt, Systems Analysis and Design, 11thed. Boston, MA: Cengage Learning, 2016.
[9] O. Jitpakde. Digital Image processing. (In Thai). Bangkok: Sakhonkit Printing & Media, 2009.
[10] “Machine Learning & Supervised, Method combination of variables.” [Online]. Available: https://media.licdn.com/dms/image/C4E12AQGPze 4iPMbjAA/articleinline_imageshrink_1000_1488/0?e=2125872000&v=beta&t=HS40LIm4NrA7nINDDthjoXlpgVbNPQUmH9wPZ1xnlaM. [Accessed: 5-Feb-2018].
[11] S. Tilley and H. J. Rosenblatt, Systems Analysis and Design, 11th ed. Boston, MA: Cengage Learning, 2016.
[12] S. Adulkasem, J. Preechasuk, and W. Adulkasem, “A System Prototype for Accurate Measurement Size of Object in X-ray Image,” (In Thai). The Journal of KMUTNB, Vol. 22, No. 1, pp. 90-98, 2012.
[13] K. Tanwong, P. Suksawang, and Y. Punsawad, “Using Digital Image to Classify Phenotype of the Rice Grain Quality under Agricultural Standards Act.” in The 22nd International Computer Science and Engineering Conference (ICSEC) 2018. Chiang Mai, Thailand, pp. 79-82, 2018.
[14] K. Tanwong, P. Suksawang, and Y. Punsawad, "Development of Rice Grain Phenotype Quality Verification System using Machine Learning," EAU Heritage Journal Science and Technology, Vol. 13, No. 1, pp. 76-94, 2019.