SVOM-VT has entered the initial phase development stage, and encircled energy is its key performance index. In the development process, it is necessary to determine the encircled energy of the optical lens stage and the system stage. The image recording of a CCD detector includes two imaging processes: one is the pixel integration imaging process, in which the output signal of each pixel is proportional to the area integration of the incident light intensity on the surface of the pixel; the other is the discrete sampling process, in which the continuous graphical object is sampled discretely at the sampling interval of the center distance of the pixel. Based on the data of SVOM-VT, the effect of CCD under-sampling on the encircled energy of detection camera is characterized by simulation and test. Imaging process of CCD pairs of scattered speckles from the lens is a two-dimensional discrete sampling process, as well as the sampling process of discrete signals. This process will lead to low-frequency noise (under-sampling noise) in the sampling of high-frequency signals by CCD detectors, resulting in spectrum aliasing (low-frequency signal distortion) of image signals. Intuitively, the original image is broadened. When the sampling density is increased, this will not be the case.
For a wavefront tested by Shack-hartmann wavefront sensor, the zonal integration method is often chosen by researchers to solve the reconstruction problem. But it has shown an unacceptable result when the phase derivative data is distributed on an unconnected domain, the obtained wavefront will contain different piston error on each subdomain. Therefore, a new zonal wavefront estimation algorithm is proposed to deal with this drawback, which uses a simultaneous fit method to correct piston error of each subdomain. The validity of the algorithm is verified by a numerical simulation and experimental results.