The noise has a serious effect on the quality of the three-dimensional holographic images reconstructed. A new method is
proposed to improve the Burch code and combine it with the three-step phase shifting method to remove the noise. The
reconstructed images are with the high contrast and resolution. The three-dimensional reconstruction images are
compared with the ones of median filtering and of 20-time intensity superposition method. The performance parameters
of three methods are analyzed. The experimental results show that the zero-order light spot, conjugate image and speckle
noise are suppressed effectively. The quality of the reconstructed image is noticeably improved.
Digital holography is a kind of coherent imaging method and inevitably affected by many factors in the process of
recording. One of dominant problems is the speckle noise, which is essentially nonlinear multiplicative noise related to
signals. So it is more difficult to remove than additive noise. Due to the noise pollution, the low resolution of image
reconstructed is caused. A new solution for suppressing speckle noise in digital hologram is presented, which combines
Canny filtering algorithm with wavelet threshold denoising algorithm. Canny filter is used to obtain the edge detail.
Wavelet transformation performs denoising. In order to suppress speckle effectively and retain the image details as much
as possible, Neyman-Pearson (N-P) criterion is introduced to estimate wavelet coefficient in every scale. An improved
threshold function is proposed, whose curve is smoother. The reconstructed image is achieved by merging the denoised
image with the edge details. Experimental results and performance parameters of the proposed algorithm are discussed
and compared with other methods, which shows that the presented approach can not only effectively eliminate speckle
noise, but also retain useful signals and edge information simultaneously.