Tea Flushes Identification Based on Machine Vision for High-Quality Tea at Harvest

Abstract:

Tea flushes identification from their natural background is the first key step for the intelligent tea-picking robot. This paper focuses on the algorithms of identifying the tea flushes based on color image analysis. A tea flushes identification system was developed as a means of guidance for a robotic manipulator in the picking of high-quality tea. Firstly, several color indices, including y-c, y-m, (y-c)/(y+c) and (y-m)/(y+m) in CMY color space, S channel in HSI color space, and U channel in YUV color space, were studied and tested. These color indices enhanced and highlighted the tea flushes against their background. Afterwards, grey level image was transformed into binary image using Otsu method and then area filter was employed to eliminate small noise regions. The algorithm and identification system has been tested extensively and proven to be well adapted to the complexity of a natural environment. Experiments show that these indices were particularly effective for tea flushes identification and could be used for future tea-picking robot development.

• Periodical: Applied Mechanics and Materials (Volume 288)
• Pages: 214-218
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.288.214
• Online since: February 2013
• Authors: Xiao Jun Jin, Yong Chen, Ying Qing Guo, Yan Xia Sun, Jun Chen
• Keywords: High-Quality Tea, Machine Vision, Tea Flushes Identification, Tea-Picking Robot

References

[1] Hayashi, S.; Ganno, K. and Y. Ishii: Machine vision algorithm of eggplant recognition for robotic harvesting. J. SHITA, Vol. 12, p.38–46 (2000).

DOI: 10.2525/jshita.12.38

[2] Bulanon, D. M.; Kataoka, T.; Ota, Y.; Hiroma, T: A Segmentation Algorithm for the Automatic Recognition of Fuji Apples at Harvest. Biosystems Eng. Vol. 83, pp.405-412 (2002).

DOI: 10.1006/bioe.2002.0132

[3] Slaughter, D., D. Gilesa and D. Downeya: Autonomous robotic weed control systems: A review. Comput. Electron. Agr., Vol. 61, p.63–78 (2008).

[4] Jin Xiaojun, Chen Yong, Sun Yanxia: Research advances of weed identification in agricultural fields. Journal of Agricultural Mechanization Research. Vol. 33, pp.23-27 (2011).

[5] Woebbecke, D. M.; Meyer, G. E.; Bargen, Von, K.; Mortensen, D. A: Color indices for weed identification under various soil, residue, and lighting conditions. Trans. ASAE. Vol. 38, pp.259-269 (1995).

DOI: 10.13031/2013.27838

[6] Meyer, G. E.; Neto, J. C: Verification of color vegetation indices for automated crop imaging applications. Computers and Electronics in Agric. Vol. 63, pp.282-293 (2008).

DOI: 10.1016/j.compag.2008.03.009

[7] Gee, Ch.; Bossu, J.; Jones, G.; Truchetet, F: Crop/weed discrimination in perspective agronomic images. Computers and Electronics in Agric. Vol. 60, pp.49-59 (2008).

DOI: 10.1016/j.compag.2007.06.003

[8] Otsu, N: A threshold selection method from gray-level histogram. IEEE Trans. Syst. Man Cyber. Vol. 9, pp.62-66 (1979).

DOI: 10.1109/tsmc.1979.4310076