Vector quantization (VQ) is a simple and widely used compression technology in many applications. For image compression, VQ provides both a fixed compression ratio and maintains acceptable distortion. However, the performance of VQ still can be improved in terms of the image quality of compressed images and codebook size used for encoding and decoding. In this paper, a new VQ-like image compression method is proposed to improve the performance of traditional VQ by using locally adaptive coding concept. The experimental results confirm that the image quality of the compressed image offered by the proposed method is higher than 30 dB on average, and the number of codewords used in our codebook is less than that required by traditional VQ.