Morphology measure and 3D reconstruction of corn leaf based on machine vision
基于立体视觉的玉米叶片形态测量与三维重建

A method was proposed to gain the plant configuration information and 3D reconstruction. According to the principles of binocular stereo vision, two images were obtained. By image segmentation technology and edge detection algorithm, leaf edge of each image was extracted. Epipolar line and RGB of color image were applied to match stereo images, then 3D coordinates of the leaf edge were worked out, so the 3D edge of the leaf could be reconstructed. Besides, using the same technique, the surface of the leaf could be reconstructed. According to the point cloud, the distance between any two points could be got. Results of many experiments show that 3D information of corn leaf can easily be reconstructed by this method.