The invention discloses a
remote sensing image generalized shadow spectrum
reconstruction method and
system based on statistics, and the method comprises the steps: obtaining
remote sensing image datato be subjected to spectrum reconstruction, carrying out the
radiometric calibration of the
remote sensing image data, converting the remote sensing image data into
radiation data, and carrying out the data
cutting of a to-be-subjected spectrum reconstruction region; performing
mixed spectrum decomposition on the spectrum reconstruction area after data
cutting, and extracting a shadow base component; deducing a generalized shadow consistency spectrum reconstruction equation according to a
radiation transmission theory; and based on statistical equation
estimation, obtaining spectral parameters, and realizing generalized shadow
spectral reconstruction of the remote sensing image to be subjected to
spectral reconstruction. According to the method, the classical definition of traditional remote sensing image shadows is broken through, a generalized shadow consistency spectrum reconstruction equation is derived, the reconstruction result is natural in
color transition, the texture is moreprominent, no edge effect exists, and the effect is very remarkable; and the fidelity of the information recorded by the image data is kept to the greatest extent.