论文部分内容阅读
The research purpose of this paper is to show the limitations of the existing radiometric normalization approaches and their disadvantages in change detection of artificial objects by comparing the existing approaches,on the basis of which a preprocessing approach to radiometric consistency,based on wavelet transform and a spatial low-pass filter,has been devised.This approach first separates the high frequency information and low frequency information by wavelet transform.Then,the processing of relative radiometric consistency based on a low-pass filter is conducted on the low frequency parts.After processing,an inverse wavelet transform is conducted to obtain the results image.The experimental results show that this approach can substantially reduce the influence on change detection of linear or nonlinear radiometric differences in multi-temporal images.
The research purpose of this paper is to show the limitations of the existing radiometric normalization approaches and their disadvantages in changing detection of artificial objects by comparing the existing approaches, on the basis of which a preprocessing approach to radiometric consistency, based on the wavelet transform and a spatial low-pass filter, has been devised. This approach first separates the high frequency information and low frequency information by wavelet transform. Chen, the processing of relative radiometric consistency based on a low-pass filter is conducted on the low frequency parts. After processing, an inverse wavelet transform is conducted to obtain the results image. The experimental results show that this approach can substantially reduce the influence on change detection of linear or nonlinear radiometric differences in multi-temporal images.