Mastication is one of the most important indexes to measure the taste quality of citrus. Most sensory evaluations are time-consuming and laborious at present. Some research is also related to the mastication of citrus that directly quantify sensory using texture characteristics. Particularly, there is the impact of citrus quality indicators on mastication properties. However, it is still lacking on the varieties and accuracy of models to quantify the mastication of citrus fruits. In this study, the convenient and accurate evaluation was proposed for the mastication of multi-batch citrus samples. 12 types of citrus fruits were selected to combine with the correlation analysis. The sensory evaluation was also used to screen the cell wall components, ultrastructure and thickness of segment membrane, in order to determine the quantitative indicators of mastication. Some indicators were then obtained for the quantitative evaluation of mastication using texture testing. Single-factor experiments and response surface analysis were carried out, where the factors were taken as the occlusal shearing with different tool types, blades thicknesses, shear rates and the number of shear petals. A combination of tool and test was selected for the optimal parameters. A regression model was then established to increase the number of citrus samples by 26. The mastication was finally classified via discriminant analysis. The results showed that the sensory scores were significantly negatively correlated with four indicators, including cellulose content, the thickness of segment membrane, number of capsule cell layers and original pectin content (P <0.01). There was a significant negative correlation between hemicellulose content and sensory average (P <0.05). While there was no significant correlation between water-soluble pectin and sensory average. Among them, a quantitative evaluation index of mastication was taken as the highest correlation coefficient (-0.92) with the thickness of segment membrane. The correlation analysis between various texture indicators and the thickness of segment membrane showed that the maximum shear force and shear work were significantly positively correlated with the capsule thickness, while the puncture force, the maximum shear stress, hardness, and adhesiveness were significantly positively correlated with the thickness of segment membrane. There was no significant correlation between chewability and the thickness of segment membrane. The maximum shear force shared the highest correlation coefficient with the thickness of segment membrane, which was 0.83 as a detection indicator; According to the single-factor experiments, the correlation coefficients between the thickness of segment membrane and the maximum shear force were basically presented a trend of first increasing and then decreasing, as the blade thickness, the shear rate, and the number of petals increased; When the correlation coefficient between the maximum shear force and the thickness of segment membrane was the highest (0.942), the blade thickness was 2 mm, the shear rate was 180 mm/min, the number of shear petals was 2, and the tool was a flat cutting double blade. The regression equation was obtained for the maximum shear force-thickness of segment membrane model, R²=0.951. The mastication index was divided into six grades in a range of 0-1, where the higher the index was, the better the mastication was; The verification test was conducted to classify the citrus mastication. 35 citrus fruits were randomly selected from the samples of the first and second batches. The prediction accuracy of improved model was 91.43%, indicating the high accuracy for the requirements. A quantitative evaluation index was obtained for the mastication and the detection on citrus fruits using texture characteristics. A set of quantitative evaluation and grading standards were achieved for the citrus mastication, according to the optimal parameters. The finding can also provide a strong reference to accurately and conveniently evaluate citrus mastication.

Human-robot collaboration is a promising means to promote orchard intelligence and reduce the over-reliance on manual work for complex agronomic practices such as fruit tree pruning, flower and fruit thinning, and harvesting. Accurate target detection and recognition of robots on humans are the basis and prerequisite for subsequent autonomous human-robot collaboration. In this study, detection and recognition of following robots for human torso were carried out in a standardized hilly orchard. A LiDAR-based human torso detection method was proposed based on the actual orchard environment. Breakpoint detection was used to cluster and segment the point clouds, and the segmentation thresholds were determined based on experimental results. The geometric attributes of the human torso were trained in the classification detection model, resulting in the extraction of six geometric attributes of the human torso. The classification model was then trained with various combinations to obtain the optimal feature combination [girth-depth-average curvature (G-D-k)] for human torso recognition in an orchard environment. Practical experiments were carried out to validate the feasibility and accuracy of the G-D-k feature combination. The experimental results demonstrate that the G-D-k feature combination can accurately recognize human bodies in orchards. The LiDAR-based detection method can achieve relatively accurate human detection and recognition in complex orchard environments, providing a reference for target detection in human-robot collaboration in orchards.