Image features optimizing for content-based image retrieval

Developing low-dimensional semantics-sensitive features is crucial for content-based image retrieval (CBIR). In this paper, we present a method called M2CLDA (merging 2-class linear discriminant analysis) to capture low-dimensional optimal discriminative features in the projection space. M2CLDA calculates discriminant vectors with respect to each class in the one-vs-all classification scenario and then merges all the discriminant vectors to form a projection matrix. The dimensionality of the M2CLDA space fits in with the number of classes involved. Moreover, when a new class is added, the new M2CLDA space can be approximated by only calculating a new discriminant vector for the new class. The features in the M2CLDA space have better semantic discrimination than those in traditional LDA space. Our experiments show that the proposed approach improves the performance of image retrieval and image classification dramatically.