Off-axis system misalignment detection technique based on hybrid features of the wavelet transform and the Gabor filter

Optical system misalignment aberrations caused by various factors such as environmental conditions and mechanical jitter of optical components can lead to a decline in the imaging quality of the system. The non-rotational symmetry structure and the coupling effect of different degrees of freedom of off-axis reflective telescopes make them more sensitive to misalignment aberrations. Moreover, when the <i>Z</i>-axis of the secondary mirror in an off-axis system has eccentricity errors, using a defocused PSF to solve the misalignment will result in a decrease in the accuracy of the lateral misalignment solution. Therefore, this paper proposes a misalignment error detection technology based on the wavelet transform and the Gabor filter and establishes a nonlinear mapping relationship between the wavelet transform-Gabor filter hybrid features of defocused PSF images and the misalignment of the secondary mirror through ShuffleNetV2. Using the hybrid features extracted by the biorthogonal wavelet transform and the Gabor filter as the input of the neural network can reduce the difficulty of decoupling the aberration features of defocused PSF images, thereby improving the accuracy of solving the misalignment of the secondary mirror. The algorithm was verified through an actual off-axis two-mirror afocal system, proving its feasibility and effectiveness.