Abstract: Considering the irregularity of seismic data and the limitation of traditional enhancing resolution method in single-channel seismic data, the problem of enhancing resolution is introduced into Curvelet domain and the method for enhancing resolution of irregular seismic data in curvelet domain is proposed. It utilizes the two-dimensional or high-dimensional support of curvelet basis function and the sparse representation of seismic data in curvelet domain, the problem of enhancing resolution of two-dimensional or high-dimensional seismic data is converted into the sparse promote solving with the constraint of L1-norm. Avoiding the problem of spatial consistency of the traditional enhancing resolution method, this method can enhancing the seismic resolution, suppressing random noise and restore the missing seismic traces, the completeness and resolution of seismic data are improved. Model and real data verify the correctness, validity and applicability of the method.