Automatic mapping aquaculture in coastal zone from TM imagery with OBIA approach

Aquaculture area monitoring is of great importance for coastal zone sustainable management and planning. This paper focuses on the development and assessment of an automatic approach for aquaculture mapping in coastal zone from TM imagery. The contribution mainly consists of three aspects: first, utilizes the Multi-scale segmentation/object relationship modeling (MSS/ORM) strategy on the object based image analysis (OBIA) of TM imagery; second, evaluates the effectiveness GLCM homogeneity texture feature on pond aquaculture area information extraction; third, compares the analysis results from three different approaches, namely pixel-based maximum likelihood classifier (MLC), One-step supervised OBIA with stand nearest neighbor (SNN) and MSS/ORM OBIA strategy. The final result shows that the MSS/ORM OBIA approach greatly improves the classification accuracy and has good potential for automatic pond aquaculture land mapping in coastal zone from TM imagery.