Contributing to the realm of information security by ensuring the secure and confidential transmission of data is the role of Reversible Data Hiding (RDH) technology. Interpolation based RDH algorithms are popular for their embedding ability and visualization. Existing methods usually fail to recover the original image completely, and this study solves this problem by optimizing the down-sampling method before the interpolation step to achieve complete recovery of the original image. Within this research, a cutting-edge RDH scheme has been introduced, featuring Optimal Neighborhood Mean Interpolation (ONMI) and the implementation of a dual-layer embedding strategy. The first layer uses OPAP and the second layer uses histogram shifting to improve the embedding capability while preserving high visual quality. The results of the experiments indicate a substantial enhancement in embedding capacity and PSNR with the suggested approach. , surpassing the performance of current RDH methods and achieving an embedding rate of 3.10 bits per pixel.