切换滤波方法在深度图像噪声处理中的应用
Switching filtering method with its application of depth image noise processing

基于Kinect深度图像的互动投影系统受到多种混合噪声干扰。单一的滤波方法只针对特定噪声，往往难以取得很好的效果，通常要将几种滤波方法联合使用。一般的联合滤波虽然可以保证图像的精确度，但算法复杂度高、耗时较长，影响了互动投影系统的实时性。针对上述矛盾，本文首次将切换系统方法引入到数字图像滤波领域，提出了一种随机切换滤波算法，并将其应用到实际的互动感投影系统中。实验结果表明，切换滤波算法既能很好地去除噪声，又能保证跟踪的实时性，达到了良好的目标跟踪效果，可以广泛应用于各种图像滤波处理系统中。
The interactive projection system based on the depth image acquired by Kinect suffers from a variety of noise interferences. A single filtering method is only for specific noise, and it is difficult to obtain good results. Therefore, it is common to combine several filtering methods for filtering the mixed noise. Although the general sum filter can ensure the accuracy of the image, the algorithm has high complexity and takes a long time, which affects the real-time performance of the interactive projection system. In view of the above contradiction, the switching system method is introduced into the field of digital image filtering, an arbitrary switching filter algorithm is proposed, and it is applied on the actual interactive projection system. The experimental results show that the switching filter algorithm can not only remove the noise well, but also guarantee the real-time tracking performance, and achieve a good target tracking effect. The algorithm can be widely used in various image filtering processing systems