On-line internal quality evaluation system for the processing potatoes

摘要

Near infrared (NIR) spectroscopy has been used to measure the multiple quality attributes in agricultural product nondestructively. To produce acceptable yields of processed potato products with good textural and color, the dry matter, specific gravity and hollow heart are important quality parameters. The objective of this research was to investigate the possibility of predicting the percentage of specific gravity, and dry matter and detect the hollow heart in potato (Solanum tuberosum L.) tubers byusing NIR sensing technique as a rapid and nondestructive method.On-line internal quality evaluation system for processing potatoes using the NIR technique was developed to predict the dry matter, and specific gravity, and classify the hollow heart. This system has NIR transmittance measuring system and a sorting system. The potatoes were fed one by one into the sorting line and they were moved to NIR transmittance measuring system which consisted of the NIR spectroscopy, specially designed light source, and CCD-array sensor. The control program of the grader designed to sort based on the weight, hollow heart, dry matter, and specific gravity information for each potato. The correlation coefficient of developed model for specific gravity was 0.87 andthe Standard Error of Correlation (SEC) was ± 0.0045. From the cross validation result, the correlation coefficient was 0.83 and the Standard Error of Prediction (SEP) was ± 0.0050. The developed model produced 83% accuracy for dry matter compared withthe measured data from the destructive drying method, and the SEC was ± 0.62%, and the correlation coefficient from validation analysis was 0.80 and the SEP was ± 0.67%. The potatoes, which have low specific gravity, can be separated, but all potatoeswith hollow heart cannot be separated. The large size of hollow heart in potato tuber can be differentiated from the all potatoes using the visible/NIR transmittance method.