Abstract: We often use the positive fuzzy rules only for image classification in the traditional image classification system. However, the role of the negative fuzzy rules also is sometimes important in image classification. Accordingly, this paper proposes a new fuzzy rule system which uses the combination of positive and negative fuzzy rules to classify images. We focus on combining negative fuzzy rules with traditional positive ones. Experiments show that the proposed method here has high accuracy and better results than the methods using positive rules only.

Key words: mage classification, positive and negative fuzzy rules, remote sensing image, natural image

摘要： 由于传统的图像分类只是利用正模糊规则对图像分类，忽略了负模糊规则在图像分类中的作用。据此本文提出用正负模糊规则的相结的方法对图像进行分类，注重将负模糊规则和传统的正模糊分类规则有效结合。实验表明，该方法有较高的准确率，获得了更好的效果。

关键词: 图像分类, 正负模糊规则, 遥感图像, 自然图像