Feasibility study of sugarcane stalks separation from trash using PCA based on color space of digital photos

12th TSAE International Conference

Panduangnate, Lalita^1 ; Saengprachathanarug, Khwantri^1^2 ; Posom, Jetsada^1^2 ; Phuphaphud, Arthit^1 ; Chea, Chanreaksa^1 ; Taira, Eizo^3

Curriculum of Agricultural Engineering, Faculty of Engineering, Khon Kaen University, Khon Kaen, Thailand^1

Applied Engineering for Important Crops of the North East Research Group, Khon Kaen University, Khon Kaen, Thailand^2

Faculty of Agriculture, Ryukyus University, Japan^3

关键词: Digital photos;    Feasibility studies;    Key quality indicators;    Labor costs;    Measurements of;    Principle component analysis;    Production efficiency;    Thresholding;   

Others  :  https://iopscience.iop.org/article/10.1088/1755-1315/301/1/012050/pdf
DOI  :  10.1088/1755-1315/301/1/012050

来源: IOP

【 摘 要 】

Percentage of trash in the harvested sugarcane is one of the key quality indicators and needed to be examined since it affects sugar production efficiency. Currently, in Thailand, measurements of trash are done by randomly sampling by labors, which leads to delay and high labor cost this article aims to study the possibility of using the Principle Component Analysis (PCA) on the digital photos to detect harvested cane out of the trash. The component values of 5 color spaces (RGB, HSV, LAB, XYZ, and YIQ) of sugarcane stalks and trash images were obtained and then analyzed using PCA. The thresholding equation of each space was determined and applied for detecting sugarcane stalk pixel. The results revealed that the quality percentage (QP) of detection of all 5 color spaces was in a range from 82.27% to 96.19%, while those of LAB and RGB were the best at 96.06% and 94.90% respectively.