Image Recognition of Rice Seeds Germinated on Panicle

摘要

A digital image analysis algorithm was developed and complemented for detection of rice seeds germinated on panicle. The rice seeds used for this study involved four hybrid rice varieties of Jinyou402, Shanyou10, Zhongyou207 and Jiayou. Sixteen morphological features and two color features were extracted from sample images. The feature of mean hue shows the strongest classification ability among all the features. Computed as the area of seed region divided by area of the smallest convex polygon that can contain the seed region, the feature of solidity is prior to the other morphological features in germinated seeds recognition. Combined with the two features of mean hue and solidity, discriminant analysis was used to classify normal rice seeds and seeds germinated on panicle. Results show that the algorithm achieved an overall average accuracy of 98.4% for both of normal seeds and germinated seeds in all varieties. The combination of mean hue and solidity was proved to be good indicator of germinated seeds. The simple discriminant algorithm using just two features show high accuracy and good adaptability.