改进小波包分析在雷达图像消噪中的应用

针对VDR雷达图像在采集、数字化和传输过程中常受到各种噪声的干扰，不利于对图像进行分析、观察和压缩的问题，提出采用改进小波包分析算法进行图像消噪处理，小波包分析具有比小波更精确的局部分析能力，但由于小波包分解的隔点采样会产生严重的频带混叠现象，文中对小波包分析算法进行改进，利用信号的频移特性，将信号进行移频处理，消除频带混叠现象，达到高质量去噪的目的；通过VDR雷达图像消噪实验证明，上述改进算法好于小波包消噪方法，更优于小波消噪方法，具有更为广泛的应用价值。

Application of Denoising of VDR Radar Image Based on Improved Wavelet Packet Analysis

Aiming at the problem that VDR radar image usually includes a great deal of noise because it is often disturbed by noise during its collecting, digitizing and delivering and it is inconvenient for image to be analyzed, observed and compressed. The improved wavelet packet analysis is presented to the denoising of VDR radar image. The wavelet packet has more accurate local analysis ability than that of wavelet analysis. However, because of down sampling of wavelet packet decomposition, the details dissatisfy the sampling theorem, and bring the problem of band overlapping . By using the characteristic of signal frequency shift, the signal frequency is shifted and the highest frequency of the signal is cut down to solve the problem of band overlapping. The image can be denoised with high quality. By simulation, the result is superior to wavelet transform and the value of its application is wider.