Abstract: To address the difficulties in evaluating pile hole quality due to the low readability and poor visualization of pile hole morphological point cloud data, this study optimizes point cloud data visualization using radius outlier removal (ROR) filtering, Levenberg–Marquardt (LM) profile circle fitting, radial rendering, and surface reconstruction techniques. The method reduces data noise through neighborhood division and threshold detection, and determines pile hole radial change information based on interlayer discretization and two-dimensional projection plane profile circle fitting. Radial rendering is used to generate intuitive and readable images of pile hole morphology, and surface reconstruction methods are employed to enhance image smoothness and obtain three-dimensional models of pile hole morphology. Experimental results demonstrate that this technique enhances the visual perception of pile hole morphology, clearly identifies the magnitude and direction of hole center deviation, and accurately locates pile hole diameter reduction and expansion.