Based on continuous wavelet transform (CWT) in three-dimensional surface reconstruction, the result of reconstruction often depends on the choice of wavelet function. The performance of wavelet function′s spatial and frequency domain determines the efficiency and effectiveness of data process. Different wavelet function itself has different performance, which usually exits different sidelobes in frequency domain. This paper discusses the impact of complex Gauss series wavelets′ frequency sidelobes in three-dimensional shape reconstruction. The calculation results show that frequency sidelobes have a smaller trend with the increase order of complex Gauss wavelet function differential.Meanwhile, theoretical analysis and experiments give the same results in local analysis abilities. In order to illustrate the frequency sidelobe question, complex Morlet wavelet and Mexican hat wavelet function are used to verify this regularity. The experimental results agree well with the theoretical analysis.