A Review Based Study of Classification of X-Ray Images Using Content Based Image Retrieval (CBIR)

Content-based image retrieval (CBIR) is an application of computer vision techniques for the image retrieval problem, specifically the search for specific medical images in large databases. In this work the X-ray images are taken from the medical database includes skull, spine, chest and leg x-ray images. The X-ray images are taken in the .jpeg format. In this project, a complete solution for search and retrieval of rich multimedia content over large databases is presented. This work initially performs preprocessing to reduce the distortions for further processing. After the segmentation process the Support Vector Machine (SVM) is used for classification purpose of X-ray images. The framework proposed in this work involves the feature vector (FV) creation and similar image retrieval (IR). Here the Invariant Moments (IMs) are used to extract the shape features and GLCM is used to extract the texture features for. These extracted features are used for the construction of feature vector. Based on the feature vector the relevant images are extracted by giving the query image. Moreover, CBIR is an effective technique, which is appropriate for large-scale indexing, is adopted, extended and integrated to the proposed framework so as to achieve optimized search and retrieval of rich media content even from large database.