High-Frequency Subpixel Vibration Measurement Based on Compressed Sensing

High-speed camera measurement technology is widely employed in sound recovery, contactless structural monitoring, fault diagnosis, digital soil mapping and other fields. High-speed image sequences are utilized to recover minute high-frequency signals that would otherwise be ignored. However, high-speed image sequence also imposes significant pressure on data transmission and processing. In certain scenarios, achieving high-speed camera measurement with the limited bandwidth is nearly impossible. In this article, high-frequency sub-pixel vibration measurement based on compressed sensing (HSVM-CS) is introduced, which utilizes sparse samplings from the image sequence to extract vibration signals and reconstruct the complete high-frequency vibration signal. Through both simulation and experimentation, we demonstrate our ability to recover the original high frequency signal ranging from 0 to 499 Hz at a sampling ratio of 0.3 in both the X and Y directions. The peak signal-to-noise ratio of the recovered 3D signals reaches 21 dB. Furthermore, the adaptive sampling strategy based on signal entropy is employed to allocate samplings reasonably, thereby enhancing the accuracy of recovery signals at the consistent sampling rate.