Application Of Remote Sensing Image Processing In City Greenland Extraction

Urban green space refers to the area covered by green vegetation or natural landscape in a city.It plays the role of purifying the city air,improving the water body of the city,regulating the humidity of the city and beautifying the urban environment.But the performance pattern of the ground is becoming more and more complex and diverse,which brings greater challenges to the statistics,processing and analysis of large amounts of data in urban planning.The remote sensing images regards as objects and the urban green space extraction regards as a target in this paper to conduct in-depth research,and obtains ground cover type,structure and area in urban areas more quickly,more definitely and detailedly.Provide more reliable data resources for city monitoring and construction.The research focused on fuzzy clustering as the main technical means,aiming at the problems of high-resolution remote sensing images with many noises,large data volume and slow processing speed,the remote sensing image obtained was converted into color space and processed in blocks.A hybrid clustering algorithm based on EM and KFCM was designed to be applied to remote sensing image segmentation in urban green space.Introduce the inertia weight to the initial clustering center of the EM algorithm to promote the global search to avoid falling into the local minimum,and introduce the neighborhood information into the nuclear fuzzy clustering to improve the ability to suppress the noise.Through the comparison of ICA,PCA,KPCA,and KICA change detection algorithms,the KICA change detection method was selected and its improvement was used in the detection of urban green space information changes.EM-KFCM clustering was performed on the informationcontaining the change components to improve the detection accuracy.Finally,the speed and accuracy of the hybrid clustering algorithm based on EM and KFCM are evaluated by the number of iterations,the image segmentation quantitative experiment criteria and the data selection metric F-measure method.The pixel-level evaluation method was selected to compare and analyze ICA,KPCA,PCA and improved KICA change detection algorithm.Taking Hanzhong City in southern Shaanxi Province as the research area,the remote sensing data of Yun-8 and Yun-8C aerial photography are used to systematically analyze,extract and detect urban green space in this region.The experimental results show that the method proposed in the paper can extract green space information accurately,detect the change of green space,improve the ability to suppress noise,and enhance the convergence speed of the algorithm,which has a certain use value.