A sectorized object matching approach for breast magnetic resonance image similarity study
Byung K. Jung, Wei Wang, Zhe Li, Seong‐Ho Son, Jung Yeop Kim · 2012
In this paper, we propose a new image retrieval method consisting of shape feature data. In this approach we assume the images are classified into single objects through other known classification methods such as K-means and SVM algorithms. From collected binary object images, we develop a new algorithm that has less computation but equal efficiency as using shape feature - the curvature of the contour. We have experimented with classified binary object image from actual breast medical images used in real medical diagnosis. Actual experimental results show that the proposed algorithm achieves equal results against traditional image retrieval using curvature of the contour with higher efficiency.