基于提升小波的图像脉冲噪声抑制方法

Approach of Image Impulse Noise Reduction Based on Li fting Wavelet

分析利用小波进行图像去噪的方法和特点 ,根据电视跟踪测量系统的实际需要提出利用易于硬件实现的提升小波滤波器对实时图像进行去噪 ,重点研究受脉冲干扰的实时图像的噪声抑制问题。小波分解的高频部分在噪声附近具有较大值 ,通过调整提升项设计提升小波滤波器 ,所设计的滤波器包含噪声特性 ,应用这些滤波器可以检测到脉冲噪声 ,利用小波的重构公式消去图像中的噪声。由于提升小波滤波器比普通小波滤波器运算量大大减少 ,因此算法易于硬件实现。

The method and characteristic of image denoising using w avelet were analysed. According to the demand of TV tracking system , an approa ch of noise reduction from real-time image using the lifting scheme wavelet that can be implemented on hardware easily was presented. The emphasis was impulse n oise reduction. High frequency components obtained by wavelet decomposition was large around impulse noise .Lifting wavelet filters were designed by changing th eir free parameters.The designed filters had features of impulse noise. Detectio n of impulse noise can be done by applying the learnt filters to noising images. Reduction of impulse noise from the image was carried out using a wavelet reco nstruction formula. Because the computation of lifting scheme filter was decreas ed largely than general wavelet filters, the algorithm was implemented easily o n hardware. There would be a perfect perspective if the lifting wavelet denoisin g method were combined with the modern signal processing devices DSP and FPGA.