摘要
针对Kinect相机采集到的深度图像存在大面积空洞及噪声的问题,提出了一种基于彩色聚类图像引导的深度修复算法。该算法使用同步采集到的Kinect彩色图像与深度图像作为输入,首先使用模糊C均值聚类算法分割彩色图像,分割后的图像作为引导图像;然后利用此引导图像分区域修复深度图像上的空洞;最后利用改进的中值滤波算法在保持边缘的同时去除深度图像中孤立的噪声点。实验表明该算法能够有效地修复Kinect深度图像中的空洞,并且能保持边缘清晰完整;从对比实验可知,对比目前的联合双边滤波算法、快速行进法来说,提出的算法在均方根误差和峰值信噪比评价指标上均有明显提高。提出的算法能够修复深度图像中的空洞和噪声,该算法是有效的。
Considering the problem that the depth image collected by the Kinect camera has large areas of holes and noise,this paper proposes a depth repair algorithm based on color clustering image guidance.The algorithm uses the Kinect color image and the depth image collected simultaneously as input.First,the fuzzy C-means clustering algorithm is used to segment the color image,and the segmented image is used as the guide image.Then the guide image is used to repair the holes in the depth image by region.Finally,an improved median filtering algorithm is used to remove the isolated noise points in the depth image while maintaining the edges.Experiments show that the algorithm can effectively repair the holes in the Kinect depth image,and can keep the edges clear and complete.From the comparison experiment,we can see that compared with the current joint bilateral filtering algorithm and fast marching method,RMSE and PSNR of the proposed algorithm have improved significantly.The proposed algorithm can repair the holes and noise in the depth image so that the algorithm is effective.
作者
常丽园
徐瑞超
CHANG Liyuan;XU Ruichao(Shaanxi National Defense Industry Vocational and Technical College,Xi’an 710300,China)
出处
《东莞理工学院学报》
2020年第5期32-37,共6页
Journal of Dongguan University of Technology
关键词
图像修复
模糊C均值
空洞修复
噪声滤波
image repair
fuzzy C means algorithm
hole inpainting
noise filtering
