Land Cover Change Detection Using Autocorrelation Analysis on MODIS Time-Series Data: Detection of New Human Settlements in the Gauteng Province of South Africa

Abstract

Abstract-Human settlement expansion is one of the most pervasive forms of land cover change in the Gauteng province of South Africa. A method for detecting new settlement developments in areas that are typically covered by natural vegetation using 500 m MODIS time-series satellite data is proposed. The method is a per pixel change alarm that uses the temporal autocorrelation to infer a change index which yields a change or no-change decision after thresholding. Simulated change data was generated and used to determine a threshold during an off-line optimization phase. After optimization the method was evaluated on examples of known land cover change in the study area and experimental results indicate a 92% change detection accuracy with a 15% false alarm rate. The method shows good performance when compared to a traditional NDVI differencing method that achieved a 75% change detection accuracy with a 24% false alarm rate for the same study area.

Rapid detection of new and expanding human settlements in the Limpopo province of South Africa using a spatio-temporal change detection method

“…An example of this is shown in figure 3, where MODIS band 1, 3 and 4 time-series is shown for a change and no-change pixel together with their corresponding ACF. The change metric was defined in [10] as the temporal correlation at a specific lag (τ ) given as…”

“…An autocorrelation function (ACF) change detection method was recently demonstrated to detect the development of new human settlements in South Africa [10]. This method uses MODIS time-series data, which have previously been shown to be separable (distinguishable) for the natural vegetation and settlement land cover classes considered in this study [11].…”

“…In the original formulation of the ACF approach [10], a single pixel's entire time-series for a single band (spanning eight years) was used as input. A change metric was then calculated by considering the properties of the ACF of the time-series.…”

“…In this paper, the aforementioned method is extended to incorporate spatial information by not only considering the change index for a single pixel but also that of its surrounding pixels to determine whether a change should be declared. The proposed Spatio Temporal ACF change detection (STACD) method uses temporal only ACF change metrics as calculated using the approach in [10] on a per pixel basis and compares this temporal ACF change index with that of pixels in its neighborhood to increase performance. It is important to note that, the method is adapted to be able to easily incorporate multiple MODIS bands.…”

“…It is important to note that, the method is adapted to be able to easily incorporate multiple MODIS bands. The idea behind the Spatio-Temporal extension to the classic ACF change detection method is that when a single threshold is used over a large area (as was done in [10]) there will inevitably be areas that are more nonstationary due to, for example, drought in arid areas, large scale commercial agriculture, etc. This will result in large areas showing up as change in regional maps that are not necessarily the type of anthropogenic changes that are of interest.…”