A Spatial Visual Words of Discrete Image Scene for Indoor Localization

One of the fundamental problems in accurate indoor place recognition is the presence of similar scene images in different places in the environmental space of the mobile robot, such as the presence of computer or office table in many rooms. This problem causes bewilderment and confusion among different places. To overcome this, the local features of these image scenes should be represented in more discriminate and more robust way. However to perform this, the spatial relation of the local features should be considered. This study introduces a novel approach for place recognition based on correlation degree for the entropy of covariance feature vectors. In fact, these feature vectors are being extracted from the minimum distance of SIFT grid features of the image scene and optimized K entries from the codebook which is constructed by K means. The Entropy of Covariance features (ECV) issued to represent the scene image in order to remove the confusion of similar images that are related to different places. The conclusion observed from the acquired results showed that this approach has a stable manner due to its reliability in the place recognition for the robot localization and outperforms the other approaches. Finally, the proposed ECV approach gives an intelligent way for the robot localization through the correlation of entropy covariance feature vectors for the scene images.