The efficient method for encoding the remote observation data has been considered. The description of the boundaries of the objects of interest can be used when extracting the features in the process control problem. The solution of the problem of compact presentation of the object boundary based on the static approach has been shown. In the paper, the solution of the problem of compact description of object boundaries by using uncorrelated transform coefficients is considered. The efficiency of the description is achieved by presenting the boundary as a functional series of the coefficients of expansion in the basis of eigenfunctions of the covariance matrix. The compression process is based on the function of distribution of the 2-D dispersion of the Hotelling coefficients. Application of the dispersion criterion when describing the boundary of the object under investigation makes it possible to compress the data actually without losses. In this case, the restoration error approaches zero value. The results of assessment of the efficiency of encoding the boundary image have been presented.
License Holder: ©2017 - TU Ilmenau
Use and reproduction:
All rights reserved