Based on Side-Match Vector Quantization (SMVQ), the authors propose two schemes to embed more secret information without significantly sacrificing the compression ratio. The data transform scheme improves the embedding capacity while maintaining the same compression ratio. The direct embedding scheme improves the embedding capacity at the expense of increasing the compression ratio slightly. According to various demands of a user, the integrated system can adaptively select a suitable method from the proposed ones. Experimental results show that the adaptive system provides the users with full flexibility and preserves a balance between the embedding capacity and compression ratio.