Application of Texture Analysis Algorithms to Characterization of Complex Features in Urban Transportation Environment using High-Resolution Satellite Imagery

High-resolution satellite imagery has been regarded as one of important datasets to analyze urban environment composed of complex features. In this study, texture-based image analysis was applied to characterize urban features for regional planning, and firstly new implementation with user interface and main algorithms for this purpose was performed for texture image generation and wavelet image processing. As for quantitative texture analysis in the spatial domain, extraction and imaging of texture parameters based on GLCM (Grey Level Co-occurrence Matrix) was used: Homogeneity, Dissimilarity, Energy, Entrophy, ASM (Angular Second Moment), and Contrast. Results obtained by texture analysis in the frequency domain based on wavelet image processing framework provide useful indicator to delineate complex features within urban transportation environment. Through this study, experiment cases with actual satellite imagery such as KOMPSAT-EOC (6.6 M) were carried out to demonstrate and characterize complex types of urban features for regional planning, respectively. In urban remote sensing application with highresolution imagery, it is concluded that these approaches based on integrated texture analysis in the both domain can be considered as one of practical application schemes for further multi-layer classification and recognition.