It is imminent to develop intelligent harvesting robots to alleviate the burden of rising costs of manual picking. A key problem in robotic harvesting is how to recognize tree parts efficiently without losing accuracy, thus helping the robots plan collision-free paths. This study introduces a real-time tree-part segmentation network by improving fully convolutional network with channel and spatial attention. A lightweight backbone is first deployed to extract low-level and high-level features. These features may contain redundant information in their channel and spatial dimensions, so a channel and spatial attention module is proposed to enhance informative channels and spatial locations. On this basis, a feature aggregation module is investigated to fuse the low-level details and high-level semantics to improve segmentation accuracy. A tree-part dataset with 891 RGB images is collected, and each image is manually annotated in a per-pixel fashion. Experiment results show that when using MobileNetV3-Large as the backbone, the proposed network obtained an intersection-over-union (IoU) value of 63.33 and 66.25% for the branches and fruits, respectively, and required only 2.36 billion floating point operations per second (FLOPs); when using MobileNetV3-Small as the backbone, the network achieved an IoU value of 60.62 and 61.05% for the branches and fruits, respectively, at a speed of 1.18 billion FLOPs. Such results demonstrate that the proposed network can segment the tree-parts efficiently without loss of accuracy, and thus can be applied to the harvesting robots to plan collision-free paths.

Calibration and experimental verification of discrete element parameters of Panax notoginseng root

The measurement of grapevine phenotypic parameters is crucial to quantify crop traits. However, individual differences in grape bunches pose challenges in accurately measuring their characteristic parameters. Hence, this study explores a method for estimating grape feature parameters based on point cloud information: segment the grape point cloud by filtering and region growing algorithm, and register the complete grape point cloud model by the improved iterative closest point algorithm. After estimating model phenotypic size characteristics, the grape bunch surface was reconstructed using the Poisson algorithm. Through the comparative analysis with the existing four methods (geometric model, 3D convex hull, 3D alpha-shape, and voxel-based), the estimation results of the algorithm proposed in this study are the closest to the measured parameters. Experimental data show that the coefficient of determination (R²) of the Poisson reconstruction algorithm is 0.9915, which is 0.2306 higher than the coefficient estimated by the existing alpha-shape algorithm (R² = 0.7609). Therefore, the method proposed in this study provides a strong basis for the quantification of grape traits.