Title: Anomaly detection using VCA algorithm for multi-temporal hyperspectral images
Abstract: Anomaly detection in hyperspectral images using multi-temporal data is useful in applications like crop monitoring and military surveillance. Spatial resolution of hyperspectral image is in the order of 4 × 4 m to 20 × 20 m, so one pixel may contain more than one material. The spectral signature of an image at given pixel may be mixing of spectral signatures of more than one materials. Hyperspectral unmixing is the process to determine number of materials present in mixed pixel, the spectral signatures of mixing materials and their fractional proportion. Hyperspectral unmixing enables variety of applications like anomaly detection, change detection, mineral exploitation and manmade material identification and detection, and target detection. Hyperspectral unmixing involves two steps, first estimates the spectral signature of pure material presents in image, known as endmembers and second determines their proportion in mixed pixels, known as abundances. Vertex Component Algorithm (VCA) is a fast and powerful algorithm. It determines the endmember signature with the assumption of presence of one pure pixel per endmember present in hyperspectral image. The paper discusses about the anomaly detection present in hyperspectral image using Vertex Component Analysis algorithm. Simulation results for anomaly detection using multi temporal hyperspectral data are discussed. Synthetically multi-temporal data sets are generated for synthetic data sets as well as for real cuprite image.
Publication Year: 2017
Publication Date: 2017-03-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 3
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot