In the actual complex field acquisition process

some irresistible factors usually lead to the loss of seismic traces in space

accompanied by interference wave and false frequency

and the missing seismic data will affect the accuracy of subsequent seismic data processing. Therefore

according to the compressed sensing theory

one-dimensional random undersampling is carried out for a synthetic seismic data volume only along the shot point direction

which transforms the problem of data missing reconstruction into a simpler denoising problem. In the way of cyclic processing time slice

the convex set projection algorithm based on curvelet transform is used for data interpolation reconstruction; and on this basis

the direct numerical processing of frequency slice is proposed According to the idea of reconstruction. The experimental results show that both methods can reconstruct the missing data well

but compared with the reconstruction of time slice cycle processing

the reconstruction effect of frequency slice processing directly is better

the operation time is faster

and the signal-to-noise ratio is higher. It has certain application value for industrial production.