Augment LTrP For Medical Image Retrieval (ALMIR) By Means Of Ontology Based Annotation

摘要

In this work, an innovative image indexing and retrieval algorithm using Local Tetra Pattern (LTrP) and Texture features for content-based medical image retrieval (CBMIR) has been proposed. The main objective of the proposed work is to retrieve the accurate and related medical images from the stored database that resembles the query image. Here new technique is employed for reducing the semantic gap utilizing the ontology based annotation with semantic features as well as it reduces the sensory gap by extracting the entropy mass with the low level features. For texture classification and image retrieval co-occurrence matrix is calculated which is utilized to find texture features like Energy, Entropy, Variance and Correlation. Key value which acts as the index for enhancing image retrieval is got by the combination of all these features. The experiment results show that the proposed method has 82% recall at 45% precision and 70% recall at 35% precision compared to LTrP which has 60% recall at 45% precision and 40% recall at 35% precision and LBP which has 10% recall at 45% precision and 25% recall at 35% precision.