Abstract: Perceptual hashing has been proved to be an effective solution for multimedia indexing, authentication or watermarking. Traditional perceptual hashing schemes are typical designed only for one purpose. This study presents a multipurpose image-hashing scheme based on Mean-Removed Vector Quantization (MRVQ) for both copyright protection and content authentication. The main idea is to perform MRVQ on the original image to yield two index tables, one for copyright protection and the other for content authentication. The original gray-level image is first divided into non-overlapping small blocks. The mean value for each block is calculated and quantized by the scalar quantizer to get a mean index and the quantized mean is removed from the image block to obtain the residual vector that is further quantized by the vector quantizer to obtain the residual index. All obtained mean indices constructed the mean index table and all obtained residual indices construct the residual index table. The obtained two index tables are then transformed into two intermediate binary images based on two different mapping functions, respectively. One mapping function is based on the variance of indices in a 3x3 neighborhood and the other mapping function is based on the number of indices larger than the mean of indices in a 3x3 neighborhood. Finally, the authentication mark and permuted copyright logo are respectively XOR-ed with the two intermediate binary images to obtain final authentication and protection fingerprints. Experimental results demonstrate the effectiveness of the proposed scheme.