Study on land use change detection based on Landsat data with object-oriented method

In order to explore the effectiveness and applicability of object-oriented multi-scale segmentation algorithm in land use detection, the land use detection method in Shenyang urban was studied. Based on the Landsat remote sensing image data in 1989, 2000, 2010 and 2016, eCognition platform was used to extract and classify the land use information from the remote sensing image into five land types: water area, construction land, cultivated land, forest land and other land. Through the analysis of experimental process and results, the overall accuracy of object-oriented multi-scale segmentation method was better than 90%, and the Kappa coefficient is greater than 0.9. The object-oriented classification method could meet the detection requirements of land use change classification in Shenyang urban area, with a high accuracy and a ideal effect.