Real-time detection of fruits in orchard environments is one of crucial techniques for many precision agriculture applications, including yield estimation and automatic harvesting. Due to the complex conditions, such as different growth periods and occlusion among leaves and fruits, detecting fruits in natural environments is a considerable challenge. A rapid citrus recognition method by improving the state-of-the-art You Only Look Once version 4 (YOLOv4) detector is proposed in this paper. Kinect V2 camera was used to collect RGB images of citrus trees. The Canopy algorithm and the K-Means++ algorithm were then used to automatically select the number and size of the prior frames from these RGB images. An improved YOLOv4 network structure was proposed to better detect smaller citrus under complex backgrounds. Finally, the trained network model was used for sparse training, pruning unimportant channels or network layers in the network, and fine-tuning the parameters of the pruned model to restore some of the recognition accuracy. The experimental results show that the improved YOLOv4 detector works well for detecting different growth periods of citrus in a natural environment, with an average increase in accuracy of 3.15% (from 92.89% to 96.04%). This result is superior to the original YOLOv4, YOLOv3, and Faster R-CNN. The average detection time of this model is 0.06 s per frame at 1920 × 1080 resolution. The proposed method is suitable for the rapid detection of the type and location of citrus in natural environments and can be applied to the application of citrus picking and yield evaluation in actual orchards.

The leaf is the organ that is crucial for photosynthesis and the production of nutrients in plants; as such, the number of leaves is one of the key indicators with which to describe the development and growth of a canopy. The irregular shape and distribution of the blades, as well as the effect of natural light, make the segmentation and detection process of the blades difficult. The inaccurate acquisition of plant phenotypic parameters may affect the subsequent judgment of crop growth status and crop yield. To address the challenge in counting dense and overlapped plant leaves under natural environments, we proposed an improved deep-learning-based object detection algorithm by merging a space-to-depth module, a Convolutional Block Attention Module (CBAM) and Atrous Spatial Pyramid Pooling (ASPP) into the network, and applying the smooth_L1 function to improve the loss function of object prediction. We evaluated our method on images of five different plant species collected under indoor and outdoor environments. The experimental results demonstrated that our algorithm which counts dense leaves improved average detection accuracy of 85% to 96%. Our algorithm also showed better performance in both detection accuracy and time consumption compared to other state-of-the-art object detection algorithms.

With the development of the economy, the demand for cucumber quality is quickly increasing. The aim of this study was to elucidate the role that plant density plays in leaf photosynthesis, shoot dry matter distribution, yield and quality of cucumber in different cultivation periods under greenhouse conditions. Experimental treatments based on three plant density treatments (2.25, 3.0 and 3.75 plants m⁻²) were conducted in turn during three growth and harvest periods in a year. The results showed that the changes in photosynthesis and weekly yield per unit area were different and dependent on the harvest time, which was mainly induced by temperature and radiation. Interestingly, we found that reducing plant density did not significantly affect the photosynthesis of leaves and did not decrease weekly yield per unit area and total yield. Low-density treatment had the highest weekly yield per plant and total yield per plant in the three harvest periods, the highest ratio of dry matter being allocated to fruits and the highest contents of soluble sugar, total phenols, flavonoid, soluble protein, vitamin C (Vc), chlorophyll and carotenoids in fruits. Moreover, a relatively low nitrite content was found in fruits following low-density treatment. The study indicated that low-density treatment was associated with a high quality of fruits without reducing the cucumber’s total annual yield under natural light in the greenhouse. Hence, our study suggests that properly reducing cucumber plant density to 2.25 plants m⁻² could be a practicable approach for greenhouses in Shanghai, China.

The identification of cucumber fruit is an essential procedure in automated harvesting in greenhouses. In order to enhance the identification ability of object detection models for cucumber fruit harvesting, an extended RGB image dataset (n = 801) with 3943 positive and negative labels was constructed. Firstly, twelve channels in four color spaces (RGB, YCbCr, HIS, La*b*) were compared through the ReliefF method to choose the channel with the highest weight. Secondly, the RGB image dataset was converted to the pseudo-color dataset of the chosen channel (Cr channel) to pre-train the YOLOv5s model before formal training using the RGB image dataset. Based on this method, the YOLOv5s model was enhanced by the Cr channel. The experimental results show that the cucumber fruit recognition precision of the enhanced YOLOv5s model was increased from 83.7% to 85.19%. Compared with the original YOLOv5s model, the average values of AP, F1, recall rate, and mAP were increased by 8.03%, 7%, 8.7%, and 8%, respectively. In order to verify the applicability of the pre-training method, ablation experiments were conducted on SSD, Faster R-CNN, and four YOLOv5 versions (s, l, m, x), resulting in the accuracy increasing by 1.51%, 3.09%, 1.49%, 0.63%, 3.15%, and 2.43%, respectively. The results of this study indicate that the Cr channel pre-training method is promising in enhancing cucumber fruit detection in a near-color background.

Plant photosynthesis and biomass production are associated with the amount of intercepted light, especially the light distribution inside the canopy. Three virtual canopies (n = 80, 3.25 plants/m²) were constructed based on average leaf size of the digitized plant structures: ‘small leaf’ (98.1 cm²), ‘medium leaf’ (163.0 cm²) and ‘big leaf’ (241.6 cm²). The ratios of diffuse light were set in three gradients (27.8%, 48.7%, 89.6%). The simulations of light interception were conducted under different ratios of diffuse light, before and after the normalization of incident radiation. With 226.1% more diffuse light, the result of light interception could increase by 34.4%. However, the 56.8% of reduced radiation caused by the increased proportion of diffuse light inhibited the advantage of diffuse light in terms of a 26.8% reduction in light interception. The big-leaf canopy had more mutual shading effects, but its larger leaf area intercepted 56.2% more light than the small-leaf canopy under the same light conditions. The small-leaf canopy showed higher efficiency in light penetration and higher light interception per unit of leaf area. The study implied the 3D structural model, an effective tool for quantitative analysis of the interaction between light and plant canopy structure.