A vortex detection scheme using the wavelet transform of the discriminant of the eigenvalues of the velocity gradient matrix (d 2) is presented in this paper. The use of d 2 field results in better eduction capability over the previously used test fields such as enstrophy since it automatically distinguishes shear layers from vortices. Level sets are used to refine the shape of the vortex without causing a huge computational penalty. Further, the algorithm is easily automated to aid batch processing. The detection scheme was applied to swirl flow fields, successfully estimating the vortex location, the core radius and the vortex shape.