Image Feature Extraction of FMCW Signal for Unsteady Rough Flow Interface

On the basis of Fast Fourier transform, the FMCW signals of unsteady rough flow interface are extended on a multi-time scale. This method avoids the shortcomings of poor anti-interference ability of the conventional peak-searching distance-calculating method. After the Fast Fourier transform is performed on sampling data in a fixed time period, a two-dimensional matrix containing distance information is generated. Based on the variance theory, an FMCW image feature extraction algorithm is proposed, which can achieve the accuracy of 97.5% for low SNR signal extraction even in harsh environments. At the same time, the iterative algorithm in image threshold segmentation is also applied to the distance information extraction of rough flow interface, and the useful distance information is extracted from fuzzy wideband.