High Spatial Resolution Remote Sensing Image Segmentation Adaptive Scale Calculation

Along with the development the technology of spatial detection,satellite remote sensing has a major development for earth observation,especially in respect of high spatial resolution remote sensing image(Hereinafter referred to as HSRRSI),the data of HSRRSI volume increase sharply.But the development of the corresponding processing capacity of HSRRSI data is slow,seriously restricting the HSRRSI application level.Object-based Image Analysis(OBIA)provides a new idea for HSRRSI segmentation and classification.The multi-scale segmentation of image is the key of OBIA,and the segmentation adaptive scale computation is the basis of multi-scale image segmentation,which affects the accuracy of multi-scale segmentation and the efficiency of automatic extraction of image objects directly.The existing method of calculating the scale parameters has a strong dependence on a priori experience,mostly only for specific remote sensing images,the paper summarizes the existing scale calculation method based on the advantages and disadvantages,the spatial scale calculation of the high spatial resolution remote sensing image and the range scale computation under the condition of the spatial scale constraint are present,the main research contents are as follows:(1)Analyze and summarize the application scope and advantages and disadvantages of the existing scale calculation method,and it is pointed out that the trend of the future development of scale parameter calculation is the adaptive multi-scale fine scale calculation method.(2)The image filter of HSRRSI segmentation is studied,and experiments show that the appropriate filtering method can effectively eliminate the noise information in the image and image information homogeneity of intrinsic pixel object in semantic image,the algorithm retain the image edge details well,which effectively improves the segmentation accuracy and efficiency of HSRRSI.(3)For the traditional Canny edge detection algorithm can only extract the edge of a single channel image,and according to the characteristics of HSRRSI,an improved Canny vector edge detection algorithm is proposed,and experiments show that the algorithm can effectively extract the vector edge information of HSRRSI.(4)The spatial scale calculation of the high spatial resolution remote sensing image based on vector edge and the range scale computation under the condition of the spatial scale constraint are present,and the mean-shift segmentation method is used to multi-scale segmentation of the range scale image,which verifies the rationality of the calculation of the spatial scale and the calculation of the range scale.