航空发动机精密零件的CT图像增强算法

CT Image enhancement algorithm of aeroengine precision parts

受微焦CT射线源剂量不稳定、探测和采集噪声以及射线多能谱因素影响,航空发动机精密零件CT图像难免含有噪声和伪影,故对其进行图像增强处理具有重要意义。射束硬化效应会使重建图像中出现杯状伪影,为了消除这种影响,提出了一种简单实用的校正方法。该方法首先对采集的数据取对数,然后和空气对数值作差,得到多色射束衰减系数的积分;其次设置幂函数变换曲线参数,并且进行硬化校正处理;最后进行卷积反投影重建。为了进一步降低图像中的噪声,采用各向异性扩散滤波模型对图像进行去噪滤波。试验结果表明,所提出的方法能够明显提高CT图像质量,降低图像伪影和噪声。

Due to the unstable dose of micro focus CT ray source,detection and acquisition noise and multi-energy spectrum ray,the CT image of aeroengine precision parts inevitably contain noise and artifacts.Therefore,image enhancement processing is of great significance.The beam hardening effect causes cup artifacts in the reconstructed images,in order to eliminate this effect,a simple and practical correction method was proposed.Firstly,the logarithm of the collected data was taken,and then the difference with the air lograithm value was made to obtain the integral of the multi-color beam attenuation coefficient.Secondly,the parameters of power function transformation curve was set and hardening correction was carried out.Finally,convolution back projection reconstruction was carried out.In order to further reduce the noise in the image,the anisotropic diffusion filtering model was used to denoise the image.Experimental results show that the proposed method can significantly improve the CT image quality and reduce image artifacts and noise.