Remote Sensing Images Road Extraction Based On Mathematical Morphology

With the rapidly development of space techniques and information techniques, the high resolution remote sensing satellite can supply human with the digital images of high accuracy and large capability. Target extraction from remotely sensed images has become the important means of the space information updating, and has been used extensively in the national economic production and martial target detection. Hence it is very necessary to extract targets especially line targets from remotely sensed images. This paper studies the extraction of line targets such as roads from remotely sensed images based on mathematical morphology.Mathematical morphology is an applied mathematical branch for the quantitative analysis of spatial structures efficiently based on concepts of shape and structural element. Mathematical morphology has been applied to image, signal processing and geomatics et al. for many applications.Based on mathematical morphology and high-resolution satellite remotely sensed images, this paper developed the following models and algorithms:(1) Developed an approach to strengthen the target's features based on morphological Top-hat operation and gray dilation;(2) Presented a developed approach to extract line targets especially roads automatically from high-resolution remotely sensed images for mountain areas based on a new morphological operation - removal operator and other operations.(3) Built an algorithm of morphological feature segment, and developed an approach to extract road network of urban areas from high-resolution remotely sensed images.The assessment results indicate that the proposal approach is efficient for extracting line targets from remotely sensed images. The developed method leads to a potential development direction for GIS data capture based on the high-resolution satellite images. This is very important for the GIS data updating, particularly for those urban areas where ground features change rapidly. The approaches can be applied to other fields, such as automatic mapping, extracting roadwork for car navigation and many more.