Content – based image retrieval system for radiological images

Today, because of development in imaging technology,the number of images are increasing rapidly and access to them gets difficult.Images are in a structure that we cannot describe them with words but only can be perceived intuitively, thus, description and accession of images are quite difficult. Therefore, images are represented by numerical valuesusing its textural characteristics. Thus, accessionprocess to desired image is performed with an objective approach by avoiding from subjectivity. In this thesis, a content-based retrieval system for radiological images was developed using textural features of images. With this system, access to images are provided by getting desired and similar images among radiological images. This developed method provides access to similar incidents and make to convenient in the diagnostic stage in medical field. Furthermore, it provides to make an objective assessment by avoiding from subjectivity. In this way, the use of large-sized images in databases provides benefits in terms of both time and assessment.