Medical Image Retrieval using Salient Point Detector
W. Liu, Qiqi Tong · 2005
Salient or interest points ot a medical image can represent local properties of it. In this paper, a salient points based approach for medical image retrieval is proposed. In order to extract salient points from a medical image, we proposed a novel approach based on spectrum energy variation. Moreover, the other two salient points algorithm are also used. One is entropy-based and the other is a saliency-based bottom-up visual attention computational model which is motivated by visual physiological experimental results. Preliminary comparison study of salient points-based and entire image-based retrieval approach showed that our approach can achieve retrieval accuracy closely to global approach. Furthermore, amount of computation can be reduced by our approach.