New Discrete Cosine Transformation Algorithm in Image Retrieval

摘要

To improve the efficacy of traditional medical image retrieval, this paper presents a new medical algorithm that uses discreet cosine transformation innovatively. This paper chose 42 pictures from upper torso CT images obtained from patients with hepatocellular carcinoma in ShengJing Hospital to serve as databases. Then by comparing the retrieval results, 6 pictures' similarity to the given query image, gotten by different retrieval algorithms to judge the efficacy of different algorithms. In the results, DCT algorithms display better retrieval performance than traditional grayscale histogram cumulative algorithm, extracting more similar images obviously. If scholarships overcome some drawbacks of this algorithms, this technology would have promising prospect to be applied for medical system in hospital.