Dual tree complex wavelet transform and Bayesian estimation based denoising of poission-corrupted X-ray images

Degradation of digital X-ray medical image is usually caused by effects of X-ray scattering and electrical noise etc. But the Quantum noise is the main noise source of X-ray medical images and can be modeled by a Poisson process. An algorithm for image restoration in the presence of Poisson noise is presented in this paper. The approach is based on dual tree complex wavelet transform and Bayesian estimation, the threshold function is derived using Bayesian estimation theory. Compared with the traditional discrete wavelet transform, the method based on dual tree complex wavelet transform and Bayesian estimation has the properties of approximate shift invariance and more directionality, and better results were obtained by using the proposed algorithm in terms of PSNR. The method allows fast and efficient denoising of images obtained under low X-ray radiation conditions.