Buddha Image Recognition Using SITF and SVT

Main Article Content

พีรพล คำพันธ์ อรณิช ปีแหล่ พันธุ์ธิดา ลิ้มศรีประพันธ์


This paper presents image recognition for Buddha image classification. The algorithm applies Scale Invariant Feature Transform (SIFT) and Scalable Vocabulary Trees (SVT) to classify images of Buddha Image into seven classes; Phra Buddha Chinarat class, Somdet Nang PhayaRuankaew class Phra At Tharot class, Luangpho Tong Lhaima class, Phra Sri Satsada class, Luangpho Dam class and Phra Sam Phinong class. The data set of 860 images are used in the experiment, the training set consist of 90% of total images of each class (777 images) and the testing set consist of 10% of total images of each 46 class (83 images). The proposed Buddha Image recognition algorithm using SIFT and SVT yield 90.6 % 47 of accuracy.


Article Details

How to Cite
คำพันธ์พ., ปีแหล่อ., & ลิ้มศรีประพันธ์พ. (2018). Buddha Image Recognition Using SITF and SVT. Journal of Industrial Technology Ubon Ratchathani Rajabhat University, 8(2), 93-101. Retrieved from https://www.tci-thaijo.org/index.php/jitubru/article/view/163477
Research Article


[1] Department of Tourism. Tourism Receipts from International Tourist Arrivals January-December 2015 [Internet].2016 [cited 2017 October 3]. Available from: https://www.tourism.go.th/assets/portals/1/files/Tourism%20Receipts%20from%20International%20Tourist%20Arrivals.pdf
[2] National Statistical Office.International Tourist Arrival to Thailand in 2015 [Internet].2016 [cited 2017 October 3]. Available from: https://service.nso.go.th/nso/web/statseries/statseries23.html (in Thai)
[3] Lowe DG. Distinctive image features from scale-invariant keypoints. International Journal of Computer Vision.2004; 60(2): 91–110.
[4] Boonsivanon K, Wannasin R. A Development of AR Application for Tradition Herbal Leave’s Recognition. Prae-wa Kalasin Journal of Kalasin University. 2015; Special Issue: 9-23.
[5] Zhou R, Zhong D, Han J. Fingerprint Identification Using SIFT-Based Minutia Descriptors and Improved All Descriptor-Pair Matching. Sensors. 2013;13(3) : 3142-3156.
[6] Xiong Y, Wen Y, Wang P, Lu Y. Text-independent writer identification using SIFT descriptor and contour-directional feature. Document Analysis and Recognition (ICDAR).2015 13th International Conference on; 2015 August 23-26; Tunis, Tunisia: IEEE; 2015. 91-95.
[7] TAT PHITSANULOK, Tourism Authority of Thailand. [Internet]. 2015 [cited 2017 October 3]. Available from: https://www.tourismthailand.org/fileadmin/upload_img/Multimedia/Ebrochure/846/Phitsanulok-EN-58-1535365820.pdf (in Thai)
[8] Phirasant J. Development Path to Participatory Community along the Road. Royal Land No. 12 (Phitsanulok - LomSak).Bangkok: E-Library, Thailand Research Fund. 2015. (in Thai)
[9] Intaptim N.TheImage of Buddha Phra Phuttha Chinnarath Category: Research Methodologyand Data Organization. [dissertation]. Bangkok: Silpakorn University; 2010. (in Thai)
[10] National Office of Buddhism. Pang Ham Yath [Internet].2016[cited 2017 October 3]. Available from: www.onab.go.th/articles/Attitude of the Buddha/29-Pang Ham Yath (2016) (in Thai)
[11] National Office of Buddhism. Pang Aum Bath [ Internet] . 2016 [ cited 2017 October 3] . Available from: www.onab.go.th/articles/Attitude of the Buddha/9-Pang Aum Bath(2016) (in Thai)
[12] Kulavinichai P. An Analytical Study of Buddha Images as Symbols of Belief for Buddhists. Research report. PhraNakhon Si Ayutthaya: Mahachulalongkornrajavidyalaya University; 2012.
[13] Witoon Donpornphan,Sirapat Chiewchanwattana, and Khamron Sunat. Keypoints detection with Dop-RPPRBF for SIFT feature descriptor on Low Complexity Processor. 2015. [cited 2017 October 15]. Available from: https://tar.thailis.or.th/handle/123456789/42