The abstract Solving problems related to digital representation of contour images is relevant for many applications, such as environmental monitoring using aerospace sensing technology for the Earth's surface. The method involves sequential application of three algorithms for 2-D processing of the initial data. Spatio-temporal representation of the initial data, coordinate transformation and dispersion filtering are used. At the first stage of processing, spatial data are transformed into a 2-D sequence. At the second stage of processing, the obtained data are represented by coordinate transformation coefficients. At the third stage of processing, transform filtering is implemented based on the dispersion criterion. The efficiency of describing the contour of the observed object is achieved by representing the contour as a functional series of expansion coefficients in terms of discrete eigenfunctions of the covariance matrix of spatial data. Application of the dispersion criterion to the coordinate transformation coefficients allows for virtually lossless data compression. The image restoration error in this case approaches zero. The article shows a solution to the problem of compact representation of an object contour based on statistical and transformation approaches. In this case, the efficiency is estimated from the point of view of virtually lossless data compression. The image restoration error after redundancy elimination in this case approaches zero.