At present the wavelet transform(DWT) has obtained good effect in the application of image de-noising

but DWT lacks shift-invariance and directionality. Because 2D dual-tree complex wavelet transform (DTCWT) possesses of reduced shift sensitivity and improved directionality

its de-noising ability is better than separable 2D DWT. In this paper

a threshold de-noising based on parental and neighboring coefficients is constituted and a novel DTCWT image de-noising method based on parental and neighboring coefficients (PNDTCWT) is presented. After 2D DTCWT transform of original image

shrinkage threshold of each coefficient is calculated. Then

current coefficient is de-noised with this threshold. After inverse DTCWT of these de-noised coefficients the final image is got. Experimental results show that PNDTCWT achieves better de-noising performance than other de-noising methods based on DWT and DTCWT. Its PSNRs improves 0.5dB averagely compared with parental coefficients based DTCWT de-noising method. In terms of visual quality

PNDTCWT can effectively preserve the image details when de-noising the wavelet coefficients. The profiles of objects seem smooth and confusion effect is eliminated.