Automatic identification of irrigation and drainage system in land reclamation area based on object-oriented classification

To identify irrigation and drainage system in land reclamation area automatically, an object-oriented classification method was proposed. The effectiveness of this method was compared with supervised classification method and manual screen digitization in terms of recognition accuracy and efficiency. KOMPSAT-2 high-resolution remote sensing images were selected as the experimental data, and the study area is located in Da’an city of western Jilin province. The experimental results showed that the overall recognition accuracy of object-oriented classification method was 89.64%, much higher than the accuracy of supervised classification method. More over, the object-oriented classification method is less time-consuming than manual screen digitization. The object-oriented classification method needs the least human intervention to complete the classification process and could achieve more stable results than the other two methods. Results show that the object-oriented classification is a powerful tool for remote sensing monitoring of irrigation and drainage system in land reclamation area. Meanwhile, this research provides an effective way for the identification of other ground objects in land reclamation projects.