Geothermal exploration and energy recovery involves drilling of rock formation which produces rock fragments. As the fragmented rock pieces and their assembly are uniquely correlated with the petrophysical properties of rock, the preliminary examination of shape parameters for rock fragments collected during drilling performance is attempted using high-resolution 3D x-ray computed tomographic imaging and a series of image processing techniques. The 3D stack of x-ray images is binarized to extract solid objects and individual rock fragments are independently separated by grouping connected voxels. The automated algorithms are developed to perform a series of shape parameters such as sphericity, roughness, flatness and elongation factor for each rock fragment and fragment size distribution. The distributions of gathered parameters are then subjected to the statistical evaluation to assess nonlinear fitting parameters. The feasibility of image-based estimation for rock properties based on rock fragments as a drilling by-product is discussed.