3D Visualization of Poplar Seedlings Based on ToF Camera

Abstract

Abstract. In this paper, a three-dimensional (3D) visualization method of poplar seedlings was proposed, combined with the ToF camera and digital camera. Firstly, the average distance density method was proposed and employed to process outlier of leaves. Secondly, in order to improve the speed of data fitting, the local section tangent plane normal angle method was introduced to filter redundant points and retain the necessary data, and then the control points were adopted to fit the leaf surface based on Non-Uniform Rational B-Splines (NURBS) fitting algorithm. At the same time, the trunk was fitted based on NURBS according to the control points of the trunk, and the other branches were fitted using iterative method. Finally, the three-dimensional visualization of poplar seedlings was realized. A large number of experiments were carried out, including the normal growth poplar seedlings and the water shortages poplar seedlings. The results demonstrated that the poplar seedlings visualization method proposed in this paper can quickly and accurately reflect the growth states of the poplar seedlings, especially for water-shortage detection.