This paper emphasizes on designing of a novice quantizer that is more suitable for compressing images through curvelet transform. This compression algorithm is tested on various images like plain, textured and building images. The results are compared with the existing techniques like curvelet with existing quatizer, wavelet with existing quantizer and wavelet with proposed quantzer. The proposed algorithm “curvelets with proposed quantizer” outperforms the existing techniques. The performance is evaluated through visual clarity , Peak Signal to Noise Ratio (PSNR) and compression metrics such as Compression ratio and Bit-rate.