Bag of words KAZE (BoWK) with two‐step classification for high‐resolution remote sensing images. Issue 4 (1st May 2019)

• Record Type: Journal Article
• Authors: Muhammad, Usman; Wang, Weiqiang; Hadid, Abdenour; Pervez, Shahbaz
• Is Part Of: IET computer vision. Volume 13:Issue 4(2019)
• Page Start: 395
• Page End: 403
• Publication Date: 2019-05-01
• Subjects: image matching -- geophysical image processing -- image classification -- remote sensing -- feature extraction -- image resolution -- image fusion -- image representation; BoWK descriptor -- conventional KAZE descriptor -- higher classification performance -- rich informative features -- scene images -- remote-sensing data sets -- traditional feature fusion strategies -- robust fusion approach -- effective fusion approach -- feature fusion strategy -- canonical correlation analysis -- separated feature descriptors -- multineighbourhood KAZE features -- innovative BoW KAZE descriptor -- BoW representation -- spatial pyramid matching -- image patches -- novel multineighbourhood KAZE -- BoW approach -- KAZE-based image descriptor -- remote-sensing domain -- high spatial resolution imagery -- scene categories -- inter-class similarity -- recent state-of-the-art methods -- scene classification -- bag-of-words model -- high-resolution remote sensing images -- two-step classification; Computer vision -- Periodicals; Pattern recognition systems -- Periodicals; 006.37
• DOI: 10.1049/iet-cvi.2018.5069
• Languages: English
• ISSNs: 1751-9632

Abstract

Abstract : The bag‐of‐words (BoW) model has been widely used for scene classification in recent state‐of‐the‐art methods. However, inter‐class similarity among scene categories and very high spatial resolution imagery makes its performance limited in the remote‐sensing domain. Therefore, this research presents a new KAZE‐based image descriptor that makes use of the BoW approach to substantially increase classification performance. Specifically, a novel multi‐neighbourhood KAZE is proposed for small image patches. Secondly, the spatial pyramid matching and BoW representation can be adopted to use the extracted features and make an innovative BoW KAZE (BoWK) descriptor. Third, two bags of multi‐neighbourhood KAZE features are selected in which each bag is regarded as separated feature descriptors. Next, canonical correlation analysis is introduced as a feature fusion strategy to further refine the BOWK features, which allows a more effective and robust fusion approach than the traditional feature fusion strategies. Experiments on three challenging remote‐sensing data sets show that the proposed BoWK descriptor not only surpasses the conventional KAZE descriptor but also yields significantly higher classification performance than the state‐of‐the‐art methods used now. Moreover, the proposed BoWK approach produces rich informative features to describe the scene images with low‐computational cost and a much lower dimension.