Elliptic Object Features Extraction and Measurement in Image Data Mining

In image data mining, especially in mining of high dimensional association rules, extracting features of shape, color and texture is not enough for these tasks. All kinds of the features of object are often extracted. These object features of image include visual features, statistical features, and dimensional features and so on. This paper proposes an integrated method for object features extraction to meet the expression demand of high dimensional association rules. At last, the validity has been proven by an instance.