Abstract
This goal of this study is to develop an automatic discrimination system for stock seedlings. The characteristics (size, shape and color) of seedlings are key indices of property which have been used for eightfold seedling discrimination by human eyes. An automatic eightfold seedling discrimination method based on fuzzy integration theory was developed using the characteristics that were measured by image processing. The differences between eightfold and single seedlings were larger among the characteristics in the complexity degree as well as seedling saturation. These were effective indices for eightfold seedling discrimination. The results showed that it is possible to do eightfold seedling discrimination based on fuzzy integration theory by extracting the features of the area, the complexity degree, and the saturation of the seedlings from the imege. The discrimination accuracy of the eightfold seedling discrimination was 88.4%.
