Conclusion and Outlook
摘要
This book discusses the theoretical framework and detection pattern of multi-feature change detection of high-resolution remote sensing image objects based on object-oriented, proposes an image segmentation optimisation method based on quality evaluation index and hierarchical parameter selection; designs a multi-feature selection process of image objects based on the genetic particle swarm optimisation algorithm combined with the mean variance ratio fitness function, obtains reliable multi-feature combinations of image objects; constructs a multi-feature change detection theoretical framework based on adaptive weight change analysis, and a “two-step” change detection pattern, achieves better change detection effects than conventional change detection methods, and realises the distinction and identification of multiple types of changes.