Intravital imaging of brain vasculature through the intact cranium in vivo is based on the evolution of the fluorescence intensity and provides an ability to characterize various physiological processes in the natural context of cellular resolution. The involuntary motions of the examined subjects often limit in vivo non-invasive functional optical imaging. Conventional imaging diagnostic modalities encounter serious difficulties in correction of artificial motions, associated with fast high dynamics of the intensity values in the collected image sequences, when a common reference cannot be provided. In the current report, we introduce an alternative solution based on a time-space Fourier transform method so-called K-Omega. We demonstrate that the proposed approach is effective for image stabilization of fast dynamic image sequences and can be used autonomously without supervision and assignation of a reference image.