Current single-view 3D human reconstruction methods primarily utilize the Skinned Multi-Person Linear Model (SMPL) to integrate structural priors but often neglect essential spatial and depth information. This oversight leads to issues such as fragmented or disembodied body parts and a lack of detail, arising from inadequate use of spatial data. To tackle these challenges, we introduce the Point-related Implicit Function (PrIF), a novel method for reconstructing dressed bodies using single-view RGBD input via an implicit function. Specifically, to address depth ambiguities, we suggest combining RGBD with the Pixel-aligned Implicit Function (PIFu). The RGBD-PIFu employs a fully convolutional RGBD image encoder alongside an implicit function. Within this framework, we develop an innovative point cloud-related implicit function reconstruction strategy to address problems with fragmented or incomplete body parts, incorporating point cloud feature extraction and a relative encoding strategy. Leveraging RGBD inputs allows for extensive utilization of the depth and spatial data of the point cloud, thus enhancing the sensitivity of the implicit function to spatial information. Experimental results show that our algorithm significantly surpasses state-of-the-art in single-view human reconstruction.