The side-match vector quantizer (SMVQ) is a well known method for image coding. We propose the smooth side-match vector quantizer (SSMVQ). The SSMVQ improves coding quality by making smoother the transition of the gray levels of the adjacent blocks. Moreover, the smooth overlap-match vector quantizer (SOMVQ) is proposed to improve the performance of the overlap-match vector quantizer (OMVQ). Experimental results reveal that SSMVQ and SOMVQ have the higher peak SNR (PSNR) values, better visual perception quality, and lower bit rate than SMVQ and overlap-match vector quantizer (OMVQ), respectively.