In order to rapidly and nondestructively measure the height of each vegetable seedling in plug tray grown in greenhouse, a method based on red, green, blueDepth (RGB-D) camera was proposed to extract height of each single vegetable seedling. Using Kinect fusion algorithms, RGB-D camera can create a canopy color 3D point cloud from the canopy color video stream and the depth video stream. 3D segmentation and identification of individual vegetable seedling from plug tray seedlings in the complicated natural scene was a key point to be resolved. Based on the principle of the RGB-D camera imaging, a method for calculating the height of each seedling in the plug tray was investigated. The procedure for processing topview color 3D point cloud of vegetable seedlings was proposed combining filtering and clustering for segmentation and identification of vegetable seedlings. The top view color 3D point cloud of bean sprouts were firstly filtered with the algorithm combined with the conditional removal and color clustering and statistical outlier removal to denoise the complicated natural scene points and noises. Individual seedlings were accurately segmented with the algorithm of Euclidean clustering. The results showed that the average measurement error of bean sprout seedling height was 2.30mm and the average relative error was 7.69%. This result can provide an effective reference solution for the extraction of the key growth parameters of seedlings. The proposed method could be used to quickly calculate the morphological parameters of each seedling and it was practical to use this approach for highthroughput seedling phenotyping. Compared with other stateofart segmentation methods, there was no need for this approach to create new training data and accompany annotated ground truth images.