Target Extraction and Simulation of Garden Waterscape Area Based on Remote Sensing Image
In this paper,a method of target extraction for landscape waterscape region was put forward.Firstly,the dark channel priority defogging method was adopted to enhance the remote sensing image of the landscape.According to the similarity weight,the remote sensing image was divided into the average image and the contrast image.Then,a filter was introduced to filter the above images respectively and thus to obtain the denoised images.The K-nearest neighbor was combined with the local binary to perform the rough separation of water and land on the remote sensing image.Moreover,the Support vector machine classifier was used to separate the land and water regions in remote sensing images.On this basis,the edge of the waterscape area was smoothed by morphological erosion operation.Fi-nally,the waterscape region of the garden was extracted.Experimental results show that the image is significantly en-hanced and the denoising effect is obvious under the application of the method,and the extraction accuracy for the wa-ter target is higher than before.
NETL
NSTL
万方数据
