A Review on Feature Descriptors for Remote Sensing Image Retrieval

This paper presents the review on results ofapplication of feature descriptors for remote sensing image retrieval. The circular covariance histogram and the rotation-invariant point triplets potential are explored here as a multi-scale texture descriptors. In remote sensing images, the availability of images containing a significantly higher amount of spatial and spectral details has Cover (a piece of ground) with flat stones or bricks; the way for new applications (e.g., hyper spectral target detection, compound object recognition, etc.) and new commercial products and it has enabled the use of a wider range of image analysis methods. On the other hand, the rapid accumulation of gigabytes worth of remote sensing data on a daily basis has provide robust and automated tools, which was designed for their management, search, and retrieval, as essential for their effective exploitation.