Recognition and localization of fruits are key components to achieve automated fruit picking. However, current neural-network-based fruit recognition algorithms have disadvantages such as high complexity. Traditional stereo matching algorithms also have low accuracy. To solve these problems, this study targeting greenhouse tomatoes proposed an algorithm framework based on YOLO-TomatoSeg, a lightweight tomato instance segmentation model improved from YOLOv5n-seg, and an accurate tomato localization approach using RAFT-Stereo disparity estimation and least squares point cloud fitting. First, binocular tomato images were captured using a binocular camera system. The left image was processed by YOLO-TomatoSeg to segment tomato instances and generate masks. Concurrently, RAFT-Stereo estimated image disparity for computing the original depth point cloud. Then, the point cloud was clipped by tomato masks to isolate tomato point clouds, which were further preprocessed. Finally, a least squares sphere fitting method estimated the 3D centroid co-ordinates and radii of tomatoes by fitting the tomato point clouds to spherical models. The experimental results showed that, in the tomato instance segmentation stage, the YOLO-TomatoSeg model replaced the Backbone network of YOLOv5n-seg with the building blocks of ShuffleNetV2 and incorporated an SE attention module, which reduced model complexity while improving model segmentation accuracy. Ultimately, the YOLO-TomatoSeg model achieved an AP of 99.01% with a size of only 2.52 MB, significantly outperforming mainstream instance segmentation models such as Mask R-CNN (98.30% AP) and YOLACT (96.49% AP). The model size was reduced by 68.3% compared to the original YOLOv5n-seg model. In the tomato localization stage, at the range of 280 mm to 480 mm, the average error of the tomato centroid localization was affected by occlusion and sunlight conditions. The maximum average localization error was ±5.0 mm, meeting the localization accuracy requirements of the tomato-picking robots. This study developed a lightweight tomato instance segmentation model and achieved accurate localization of tomato, which can facilitate research, development, and application of fruit-picking robots.

Tomatoes are a critical economic crop. The realization of tomato harvesting automation is of great significance in solving the labor shortage and improving the efficiency of the current harvesting operation. Accurate recognition of fruits is the key to realizing automated harvesting. Harvesting fruit at optimum ripeness ensures the highest nutrient content, flavor and market value levels, thus maximizing economic benefits. Owing to foliage and non-target fruits obstructing target fruits, as well as the alteration in color due to light, there is currently a low recognition rate and missed detection. We take the greenhouse tomato as the object of research. This paper proposes a tomato recognition model based on the improved YOLOv8 architecture to adapt to detecting tomato fruits in complex situations. First, to improve the model’s sensitivity to local features, we introduced an LSKA (Large Separable Kernel Attention) attention mechanism to aggregate feature information from different locations for better feature extraction. Secondly, to provide a higher quality upsampling effect, the ultra-lightweight and efficient dynamic upsampler Dysample (an ultra-lightweight and efficient dynamic upsampler) replaced the traditional nearest neighbor interpolation methods, which improves the overall performance of YOLOv8. Subsequently, the Inner-IoU function replaced the original CIoU loss function to hasten bounding box regression and raise model detection performance. Finally, the model test comparison was conducted on the self-built dataset, and the test results show that the mAP0.5 of the YOLOv8-Tomato model reached 99.4% and the recall rate reached 99.0%, which exceeds the original YOLOv8 model detection effect. Compared with faster R-CNN, SSD, YOLOv3-tiny, YOLOv5, and YOLOv8 models, the average accuracy is 7.5%, 11.6%, 8.6%, 3.3%, and 0.6% higher, respectively. This study demonstrates the model’s capacity to efficiently and accurately recognize tomatoes in unstructured growing environments, providing a technical reference for automated tomato harvesting.

The study aimed to measure the carotenoid (Car) and pH contents of carrots using hyperspectral imaging. A total of 300 images were collected using a hyperspectral imaging system, covering 472 wavebands from 400 to 1000 nm. Regions of interest (ROIs) were defined to extract average spectra from the hyperspectral images (HIS). We developed two models: least squares support vector machine (LS-SVM) and partial least squares regression (PLSR) to establish a quantitative analysis between the pigment amounts and spectra. The spectra and pigment contents were predicted and correlated using these models. The selection of EWs for modeling was done using the Successive Projections Algorithm (SPA), regression coefficients (RC) from PLSR models, and LS-SVM. The results demonstrated that hyperspectral imaging could effectively evaluate the internal attributes of carrot cortex and xylem. Moreover, these models accurately predicted the Car and pH contents of the carrot parts. This study provides a valuable approach for variable selection and modeling in hyperspectral imaging studies of carrots.

Agricultural robotics has emerged as a research area within robotics, with a particular focus on designing end effectors that are adapted to the physical characteristics of the target fruits. Acquiring a comprehensive understanding of the physical and mechanical properties specific to tomato fruits not only minimizes mechanical damage during grasping processes but also serves as a foundation for the optimal design of gripping components. In this study, the Syngenta Sibede variety of tomatoes was used as the experimental material. The reversible viscoelastic behavior and deformation characteristics of tomato fruits were approximated using a four-element Burgers model through creep testing. The fitting coefficients for the model exceeded 0.99. The creep parameters for the four ripening stages of tomatoes were obtained, and the correlation between the ripening stage, deformation value, and creep parameters was analyzed. Correlation analysis was performed to examine the relationships between each parameter and creep deformation, revealing significant and highly significant correlations. Inter-parameter correlations were also found to be highly significant. Puncture tests were conducted on tomatoes. The exocarp rupture force of the green-ripening stage was 9.224 ± 0.901 N, which was 53.87%, 70.63%, and 104.01% higher than that of the semi-ripening stage, early firm-ripening stage, and mid-late firm-ripening stage, respectively. This study suggests that when harvesting tomatoes at the semi-ripening stage and beyond, attention should be paid to trimming the stem. Compression experiments were conducted on tomatoes, and it was discovered that under the same ripening stage, the axial compressive rupture force of tomatoes was greater than the radial rupture force. Tomatoes exhibited anisotropic behavior. The grasping direction is axial, which can be used as the new design direction of the end-effector.

Background: Lettuce is a vegetable that is increasingly consumed globally, given its nutritional quality. Plant factories with artificial lighting can produce high-yield and high-quality plants. High plant density in these systems speeds up leaf senescence. Wasted energy and lower yield raised labor expenses are some of the bottlenecks associated with this farming system. In order to increase lettuce yields and quality in the plant factory, it is essential to develop cultivating techniques using artificial lighting. Methods: Romaine lettuce was grown under a developed “movable downward lighting combined with supplemental adjustable sideward lighting system” (C-S) and under a system without supplemental sideward lighting (N-S) in a plant factory. The effects of C-S on lettuce’s photosynthetic characteristics, plant yield, and energy consumption relative to plants grown under a system without N-S were studied. Results: Romaine lettuce growth and light energy consumption in the plant factory were both influenced favorably by supplementary adjustable sideward lighting. The number of leaves, stem diameter, fresh and dry weights, chlorophyll a and b concentration, and biochemical content (soluble sugar and protein) all increased sharply. The energy consumption was substantially higher in the N-S treatment than the C-S.

This paper addresses challenges in the application of existing colters in Chinese ecological tea plantations due to abundant straw roots and insufficient tillage depth. Aligned with the agronomic requirements of hilly eco-tea plantations, our study optimizes the structural advantages of the joint use of rotary tillage blades and double-disc colters to design an efficient trenching device. Our investigation explores the motion characteristics of a double-disc colter during deep trenching operations, in conjunction with rotary tillage blades. Employing discrete element method (DEM) simulations, this paper aims to minimize the working resistance and enhance the tillage depth stability. Single-factor experiments are conducted to determine the impact of key structural parameters on the tillage depth stability and working resistance. The optimal parameters are determined as a relative height of 80 mm to 120 mm, a 280 mm to 320 mm diameter for the double-disc colter, and a 10° to 14° angle between the two discs. The central composite design method is used to optimize the structural parameters of the double-disc colter. The results indicate that when the relative height is 82 mm, the diameter of the double-disc colter is 297 mm, and the angle between the two discs is 14°, the tillage depth stability performance reaches 91.64%. With a working resistance of merely 93.93 N, the trenching device achieves optimal operational performance under these conditions. Field validation testing shows a tillage depth stability coefficient of 92.37% and a working resistance of 104.2 N. These values deviate by 0.73% and 10.93%, respectively, from the simulation results, confirming the reliability of the simulation model. A field validation test further confirms that the operational performance of the colter aligns with the agronomic requirements of ecological tea plantations, offering valuable insights for research on trenching devices in such environments.