Automatic thresholding methods are used to detect spatio-temporal changes in the land subject to different natural and anthropogenic processes. Image segmentation plays an important role in this analysis, where urban sprawl detection take place with daylight images. However, recently some investigators have used nocturnal images in remote sensing imagery research. Such georeferenced data represent a good tool for analysis of the light pollution and urban sprawl. There are various physical processes involved in the radiative transfer of the light projected from the cities. Though, with a correct method based on background subtraction, any satellite remotely sensed nocturnal image can be useful in detecting urban sprawl. We base this work on thresholding processes of georeferenced nocturnal satellite images. We used a method combining digital classification techniques, geographic information systems and statistical analyzes. The proposed method is helpful because of a simple implementation and time saving. The pixel intensity of nocturnal images can offer a tool to calculate aspects related to electricity consumption and the efficiency of public lighting. We hope the results motivates other authors to study relationships with other social, natural and economic issues.

自动阈值方法用于检测受到不同自然和人为过程影响的土地的时空变化。图像分割在此分析中起着重要作用，在该分析中，城市蔓延检测是通过日光图像进行的。但是，最近一些研究者在夜间遥感影像研究中使用了夜间影像。这种地理参考数据是分析光污染和城市蔓延的良好工具。从城市投射的光的辐射传递涉及多种物理过程。但是，使用基于背景减法的正确方法，任何卫星遥感的夜间图像都可用于检测城市蔓延。我们基于地理参考夜间卫星图像的阈值处理过程进行这项工作。我们使用了一种结合了数字分类技术，地理信息系统和统计分析的方法。所提出的方法由于简单的实现和节省时间而很有帮助。夜间图像的像素强度可以提供一种工具来计算与电力消耗和公共照明效率有关的方面。我们希望研究结果能激励其他作者研究与其他社会，自然和经济问题的关系。