Leaf Image Retrieval Based on Many Types of Queries

Abstract

Content-Based Image Retrieval (CBIR) system considers one of the most common applications in computer science. Leaf image retrieval and classification has attention of many scientists.  In this study, a new proposed method suggested for image leaf retrieval. The proposed method includes four important stages. The first stage represents image gathering for the experiments then applying the preprocessing to remove the noise and enhance the resolution of the images. The second stage represents features extractions which are shape and color histogram for quires by example and histogram. While other quires based on text information retrieval. The third stage includes calculating the similarity between the extracted features of the query image and the feature of the stored image in the database. The fourth stage includes the experimental results. The experimental results evolve four types of query which are by shape (example), color histogram, annotation and description. From the obvious results concludes that the query by description and annotation has superior performance comparing with query by shape. On the other hand, the query by color histogram has very low performance.