Aliasing is one of the most common artifacts in 2D color Doppler echocardiography. Existing methods are approximate and the most precise of them require considerable amount of computations. In the proposed paper, we describe an algorithm that modifies only areas with submaximal velocities — areas most prone to aliasing, leaving other untouched in order to facilitate the process of analysis for the radiologist. Algorithm was tested on 10 in-vivo datasets of large animals and have shown the considerable precision and computation efficiency, which made it real-time compatible. Keywords: two-dimensional color Doppler, echocardiography, disjoint-set tree, connected components