OPTIMAL DESIGN OF URBAN LANDSCAPE ECOLOGICAL BALANCE BASED ON VIRTUAL REALITY TECHNOLOGY

The construction of ecological civilization like a raging fire started, vigorously building a beautiful city landscape can effectively promote the construction of ecological civilization in the process. To optimize the design of green space and landscape ecological balance, it is necessary to extract the characteristics of landscape data and calculate the local attributes of each landscape data point. In this study, an extraction method of data feature of urban green space landscape is proposed based on multi-scale inhomogeneous filtering. In which, the multi-scale inhomogeneous filtering is used to define gray level of pixel point with and without noise of the landscape data to achieve denoising effect of the landscape. Then, local property of each landscape data point is calculated and boundary curve contained in original shape is optimized by using statistic optimization method. Finally, the point with character discrimination parameter greater than the given threshold is defined as feature point. The simulation results show that the method can extract sharp feature point effectively and has high extraction precision.