针对数字散斑干涉(Digital Speckle Pattern Interferometry, DSPI)测量中散斑噪声干扰会导致相位提取困难、测量灵敏度降低的问题, 提出基于能量的DSPI相位图正余弦滤波法, 克服传统滤波方法不具有自适应性的不足, 提高了相位图质量。该方法根据正交小波包能量估算DSPI相位图中的噪声能量; 对相位图分别进行正余弦变换, 采用均值法平滑处理; 利用噪声能量设定滤波阈值, 实现相位图有效滤波, 文中将噪声能量与正余弦滤波结合可以自适应地进行降噪处理, 利用能量控制图像滤波效果。仿真和实验结果表明, 基于能量的正余弦滤波法较传统方法可以减少50%的循环次数。证明了本文提出的方法能够有效地对数字散斑干涉相位图进行滤波, 提高信噪比。

The speckle noise in digital speckle pattern interferometry (DSPI) results in difficulty in phase extraction and decreases measurement sensitivity. A denoising method using sine-cosine filtering and signal energy was proposed to smooth phase maps of DSPI. It exceeded traditional sine-cosine denoising method in adaptively and phase map quality. In the method, noise energy of DSPI phase map was estimated by energy of orthogonal wavelet packet. The phase map was transformed by sine-cosine method and was smoothed by average filtering. The phase map was finally denoised efficiently after the noise threshold was set by the noise energy. The combination of noise energy and sine-cosine filtering reduced noise adaptively. The image filtering effect was controlled by the noise energy. Emulational and experimental results both show that the proposed method can reduce the number of cycles by 50% compared to conventional methods. The proposed method can also filter the DSPI phase maps effectively and improve the noise ratio.