Development of Nondestructive Detection Algorithm for Internal Defects of Japanese Radish

Abstract

Internal defects always can be found in many produces such as Japanese radish. It is impossible to be detected by human eye. Nondestructive measurement is suitable technique for detecting internal defects like black heart, and air cavity after harvest time, which makes the radish root unmarketable in Japan. This study is aimed to develop the nondestructive detection algorithm for internal defects of radish by near-infrared spectroscopy. Band ratio and linear discriminant analysis were used to build the detection algorithm. In calibration set, the discrimination rate was 94.1% for normal radish, 87.5% for internal defects of radish, and overall success rate was 90.8%. In prediction set, the discrimination rate was 84.4% for normal radish, 95.5% for internal defects of radish, and overall success rate was 89.9%. The first derivative data and neural network system were also used to build the detection algorithm. In training set, the discrimination rate was 100% for normal radish, 82.5% for internal defects of radish, and overall success rate was 94.4%. In testing set, the discrimination rate was 97.8% for normal radish, 77.3% for internal defects of radish, and overall success rate was 91.0%.These results show the potential of the proposed techniques for detecting and predicting radish with internal quality.