Blue-green algae discriminant model based on normalized variance of spectrum vector

Facing to the more and more serious eutrophication and frequent outbreak of blue-green algae event in inland lake, the early warning of them has become urgent affairs. To monitor them in large scale rapidly and improve the extraction accuracy by remote sensing technology, MODIS data is a kind of ideal data source for its free getting, high spectrum resolution and temporal resolution. In this paper, a novel blue-green algae discriminant model is adopted based on normalized variance of spectrum vector. Experiments show it has good extraction accuracy.