Study On Road Extraction Method In Remote Sensing Image

摘要

A road extraction method combining radiation and topology feature is proposed. First, wavelet transform is used to remove noise and detail information which impact extraction of radiation feature. Then road and other objects with same spectrum are extracted by carrying out improved K-means clustering segmentation based on radiation feature. Next, rough road extraction is achieved by implementing area threshold segmentation based on topology feature. Finally, road skeleton is extracted through thinning. The results show that this approach is applied for road extraction.