Geophysical data have to be modeled on a regular grid for various numerical procedures. However, airborne data tend to be collected with fine spacing along traverses but with much coarser spacing between traverses. Gridding only honors flight-line data when the mesh size is close to the sample spacing; otherwise, high-frequency information is always lost, which creates aliasing artifacts. For example, linear trends at an acute angle with respect to flight lines are imaged as “bull’s-eyes,” which resemble a boudinage at line intersections. The presence of boudinage artifacts can significantly distort anomalies of interest and thus lead to incorrect interpretation of shapes or sizes of causative bodies. We evaluated a method called constrained coherence-enhancing diffusion filtering that only diffuses the image in specific areas where strong anisotropy is detected. This method was tested on synthetic and field data set. Results indicated that the method can be efficiently used to enhance linear structure in multiple local directions. The images derived from this grid, such as the vertical gradient map, are also significantly improved. The original line data are honored by the constraints applied. We also used a field data set to compare the proposed approach with the approach used when diffusion is applied uniformly in all areas, irrespective of anisotropy. The proposed method was proven to produce better results with fewer artifacts.