This study sought to measure accurately the thickness of the frontal, parietal and occipital bones of the alive human skull based on CT images. The images were treated with a series of processes by coding into a program: image segmentation and binary coding; eliminating edge interference; rectifying image deviation and clarifying boundary edge; redrawing the boundary and reference point fixing; and finally thickness measurement. The new method can measure not only the different points in one CT image but also the same point in massive CT images. The measurement results are of great practical value in the fields of anatomy, clinical medicine, biomechanics study, head injury analysis, etc.