Study to process of urban sprawl using spectral mixture analysis

The surface feature of urban areas is complex, there is always serious mixed pixel problem of remote sensing image. Traditional classification information extraction methods of urban will produce significant errors. Spectral Mixture Analysis (SMA) has brought hope for high-precision information extraction of urban, this paper studied urban expansion process of Shenyang in 1992-2007 based on SMA. We chose three scenes of Landsat TM/ETM+ remote sensing images of 1992, 2001, 2007, and obtained urban impervious surfaces (ISA) based on SMA, the expansion of urban space was divided into four directions (Northwest, Northeast, Southwest , Southeast). We analyzed the urban expansion process using the difference of ISA results of different years. The results show that ISA change area of low value is always larger than that of high value in the urban expansion process, it means the urban development is a gradual process; urban sprawl direction of Shenyang is mainly to the northwest, northeast and southwest between 1992-2001, to the northeast and southwest direction between 2001-2007. Take 1992-2007 as a whole year range, urban sprawl direction of Shenyang is mainly to the northwest, northeast and southwest, the sprawl area of southeast direction is small. The development of society that accelerating the urbanization process of Shenyang.