| Summary: | 碩士 === 國立台北護理學院 === 資訊管理研究所 === 94 === In the recent years, digital image database have grown enormously in both size and number. How to effectively manage the large image databases is more and more important. Many approaches have been proposed the method which allows the user to retrieve the image in database easily. In a Region-Based Image Retrieval method is used to retrieval images using the region of images. It has advantages whole retrieve in complex background and multi-object image. RBIR has been considered to be an extended CBIR method and widely employed in retrieving image database.
The goal of this thesis is to construct of region-based image retrieval system in pathologic slides. At first, the method proposed in this thesis utilizes the clustering algorithm to segment the image into regions, and extract the texture and color features to describe regions, respectively. Then we can retrieve the most similar images according the similarity between query image and database image. To totally 2000 images of pathology slides were used to test the accuracy and feasibility of the system. The results of experimentation demonstrate that the Fuzzy C-Means is effective in region generation, and actually effectively increase the retrieval performance. The system in our study established can provide the index of pathologic slides for the medical professionals to assist the clinical diagnosis and to be the tools of medical educational training.
|
|---|
