基于改进三维块匹配滤波的图像去噪算法

Image denoising based on improved 3D block matching filtering algorithm

为了有效滤除樱桃图像在获取过程中混杂的不同噪声,保障图像识别与机器自动采摘时良好的图像信息质量,提出一种改进三维块匹配滤波(block-matching and 3D filtering, BM3D)的图像去噪方法.首先,在三维块匹配滤波的基础估计阶段构建自适应中值滤波处理器,滤除图像中部分椒盐噪声,并改进优化硬阈值、滑窗步长及三维硬阈值等关键参数快速滤除高斯噪声;其次,在基础估计阶段与最终估计阶段之间引入中值滤波,最大限度地去除图像中剩余的混合噪声;最后,通过仿真实验验证所提算法的有效性,并对比分析改进前后算法的归一化均方误差、峰值信噪比、信噪比改善因子及结构相似性等性能.结果表明,改进的BM3D方法在保持好樱桃图像细节信息的同时,能有效去除高斯噪声和滤除大概率椒盐噪声,且随混合噪声干扰的增强,所提算法的去噪性能更佳且优于其他滤波方法.

In order to effectively filter out the different noises mixed in the process of acquiring cher-ry images,and ensure the good image information quality of image recognition and automatic pick-ing by the machine,an image denoising method to improve three-dimensional block matching and 3D filtering(BM3D)is proposed.Firstly,an adaptive median filter processor is constructed in the basic estimation stage of 3D block matching filtering to filter out some salt and pepper noise in the image,and the key parameters such as hard threshold,sliding window step size and 3D hard thresh-old are improved and optimized to quickly filter out the Gaussian noise.Secondly,median filter is introduced between the basic estimation stage and the final estimation stage to remove the remaining mixed noise in the image to the maximum extent.Finally,the effectiveness of the proposed algo-rithm is verified by simulation experiments,and the performance of the algorithm before and after the improvement is compared and ananlyzed,such as normalized mean squared error,peak signal-to-noise ratio,signal-to-noise ratio improvement factor and structural similarity.The results show that the improved BM3D method can effectively remove Gaussian noise and filter out high-probabili-ty pepper and salt noise while maintaining the detailed information of the big cherry image,and the denoising performance of the proposed algorithm is better than that of other filtering methods with the enhancement of mixed noise interference.