Denoising Method of 3-D Point Cloud Data of Plants Obtained by Kinect

Abstract:

In order to solve the difficult acquisition of plants’ 3-D point cloud data, the Kinect was adopted to collect the 3D point cloud data of corn. Compared with the usual 3D scanning equipment, Kinect can rapidly and efficiently acquire the data with lower cost. But the accuracy of data acquired by Kinect is low. It is valuable to denoise the data. According to the characteristics of the point cloud data acquired by Kinect, the data were preprocessed and smoothed. In this paper, a multi frame data fusion method was used to obtain more complete plant 3D point cloud data, and it played a role in smoothing. A denoising algorithm based on density analysis and depth data bilateral filtering methods were proposed to process the outlier noise and internal highfrequency noise. In the experiment of corn and eggplant internal high-frequency noise denoising, compared with the traditional bilateral filtering, the denoising time of the algorithm in this paper was only 2.71% and 1.78% of traditional bilateral filtering and the noise was well removed by adjusting the parameters. The experimental results show that the proposed method can easily and quickly remove the noise of different scales, while preserving the integrity of edge data. Consequently, the good 3-D point cloud data of the plant could be obtained.