Study On Land Use Remote Sensing Detection Based On OBIA Method

Land use change detection is based on observations of the same object or phenomenon at different times to determine the process of surface cover changes.It can extract and describe the characteristics of the target change over time,and Further quantitative analysis and deterministic expression of its changes.With the rapid development of remote sensing technology,land use change detection research has become a more active research direction in remote sensing technology.Traditional pixel-based change detection method cannot make full use of image information,low classification accuracy,slow speed and other limitations,but also easily lead to a large number of spatial data redundancy and waste of resources,can not be well adapted to high resolution Image characteristics.However,the object-oriented image analysis technology can generate image objects with clear physical meaning at any scale.It can use the object as the basic operation unit,can comprehensively use the remote sensing data spectral information,texture features,topological relations and other information,can better guarantee the change detection pattern integrity and effectively inhibit the "salt and pepper" noise.If the OBIA method is used to detect the change,it can break through the limitation of traditional change detection and improve the accuracy and efficiency of change detection.This paper introduces the basic concepts,research contents and research methods of land use change detection;The traditional methods of change detection are expounded,and their advantages and disadvantages are analyzed in detail.In view of the existing problems and shortcomings of the traditional change detection method,the related theories and processes of land use change detection using OBIA method are explored.Discusses the object-oriented remote sensing image analysis software eCognition(Yi Kang)to determine the band weights,scale parameter,homogeneous factors and other important parameters in multi-scale segmentation algorithm involves.On the basis of the above analysis,we analyze the differences of changing objects between the three kinds of multi-temporal image segmentation models,such as multi-temporal combination segmentation,single-temporal segmentation and multi-temporal segmentation respectively,evaluate the influence of multi-temporal image segmentation strategy on object-oriented change detection results,and select an optimal segmentation pattern from them.The results show that the multi-temporal segmentation pattern has a great influence on the shape and precision of the change object.The multi-temporal combined image segmentation pattern in the three modes has the best precision and the highest precision.The above model is applied to the land use change detection based on OBIA method,and the better test results are obtained,which proves the effectiveness of the method.