An automated quantitative study of the alignment of vertically aligned carbon nanotubes (VA-CNTs) from scanning electron microscopy (SEM) images has been demonstrated. It is based on the fact that the image gradient (directional change in intensity or color) at every pixel in the image contains the information about the anisotropy at that particular pixel i.e., the local alignment is maintained in the orthogonal direction to the gradient. Structure tensor metrics are formulated demonstrating to be able to summarize the distribution of gradient directions within the neighborhood of any pixel within a two-dimensional domain (surface). An image analysis is presented that evaluates the alignment in desired regions of interest (ROIs) in SEM images of VA-CNTs. This method has been exploited to study the alignment of two different kinds of VA-CNTs: one grown via the thermal chemical vapor deposition (T-CVD) method and the second synthesized via the plasma enhanced chemical vapor deposition (PE-CVD) method.