Urban vegetation coverage information extraction based on improved linear spectral mixture mode

In order to monitor urban vegetation coverage more accurately using remote sensing technology, a new model, which is fit for the extraction of urban vegetation coverage, is presented. In this model, TM image was decomposed by linear spectral mixture mode and then the vegetation component of TM image was amended by the high-resolution satellite image. Afterwards, an example of application was given to verify this model. Although errors and confusion exist, this method shows satisfying results with an overall accuracy of 95.1%. Compared with the traditional and improved dimidiate pixel models respectively, this model is more suitable for quantifying vegetation coverage in urban regions. Based on what have been mentioned above, the model has a very high practical value.