Subsignal technique for estimating parameters of signals using symmetric windows

Frequency-domain methods are widely used in optical metrology systems to measure shape, motion, and deformation. Central to these methods is the estimation of sinusoidal signal parameters, the accuracy of which directly determines the overall measurement accuracy. This paper presents a novel parameter estimation method for sinusoidal signals, which extends the recently proposed STEPS method to accommodate symmetric windows. Estimation errors induced by noise are mathematically derived and validated through simulations. A strategy of optimizing subsignal length selection is also proposed to enhance parameter estimation accuracy. The enhanced STEPS method is substantially more effective in suppressing the impact of harmonic distortion and noise. Both simulation and experimental results demonstrate that the enhanced STEPS method significantly outperforms other methods in measurement accuracy.