3D laser scanning technology, while effective for acquiring point cloud data, often encounters limitations due to factors such as the measured object, measuring instrument, and environment, leading to incomplete data and holes within the dataset. Utilizing such incomplete data directly can compromise the quality and precision of subsequent modeling and reconstruction processes. Consequently, hole repair is a critical step. This paper presents an enhanced radial basis function (RBF) algorithm for repairing point cloud data, which improves upon the accuracy and speed of the original method, making it particularly suitable for repairing convex models. This algorithm implemented in Visaual Studio 2019, repairs point cloud holes by detecting hole boundaries, initializing grids, least squares grids, establishing implicit surfaces, and adjusting points to implicit surfaces. Experiments show that this method can fill the holes on the surface well, and the repair position remains smooth with the original area, which can be applied to the repair of indoor irregular objects, caves and tunnels.