Abstract: To ensure the accuracy of performance parameter measurements and reduce the cost and complexity of low-frequency shakers, a machine vision-based method for determining the parameters of low-frequency shakers is proposed. First, the guide rail is modeled as a large-radius arc, and the displacement change of the feature marker in the vertical direction is analyzed. The bending condition is then fitted using the least squares method to measure the bending degree of the guide rail. Next, the subpixel-level edge extraction method is applied to the region of interest in the motion sequence images to accurately measure the motion displacement of the shaker table. Finally, the sine approximation method is used to fit the motion displacement of the shaker to obtain its fitting amplitude, which is then used to solve the key performance parameters of the shaker. Furthermore, high-precision measurements of these parameters can be achieved using a simple set of visual measuring devices. Comparative experimental results with traditional measurement methods show that the bending degree obtained by the machine vision method is highly similar to that of the traditional method when three different loads are added. For the measurement of other performance parameters, the machine vision method can also obtain reliable measurement accuracy and efficiency in the range of 0.01-10 Hz.