Accurate and reliable fruit detection in the orchard environment is an important step for yield estimation and robotic harvesting. However, the existing detection methods often target large and relatively sparse fruits, but they cannot provide a good solution for small and densely distributed fruits. This paper proposes a YOLOv3-Litchi model based on YOLOv3 to detect densely distributed litchi fruits in large visual scenes. We adjusted the prediction scale and reduced the network layer to improve the detection ability of small and dense litchi fruits and ensure the detection speed. From flowering to 50 days after maturity, we collected a total of 266 images, including 16,000 fruits, and then used them to construct the litchi dataset. Then, the k-means++ algorithm is used to cluster the bounding boxes in the labeled data to determine the priori box size suitable for litchi detection. We trained an improved YOLOv3-Litchi model, tested its litchi detection performance, and compared YOLOv3-Litchi with YOLOv2, YOLOv3, and Faster R-CNN on the actual detection effect of litchi and used the F1 value and the average detection time as the assessed value. The test results show that the F1 of YOLOv3-Litchi is higher than that of YOLOv2 algorithm 0.1, higher than that of YOLOv3 algorithm 0.08, and higher than that of Faster R-CNN algorithm 0.05; the average detection time of YOLOv3-Litchi is 29.44 ms faster than that of YOLOv2 algorithm, 19.56 ms faster than that of YOLOv3 algorithm ms, and 607.06 ms faster than that of Faster R-CNN algorithm. And the detection speed of the improved model is faster. The proposed model remits optimal detection performance for small and dense fruits. The work presented here may provide a reference for further study on fruit-detection methods in natural environments.

To address the automatic detection of dense and small-scale fruit targets under natural large-scene conditions, litchi was used as the research object. Here, a method to automatically detect dense and small-scale litchi fruit targets based on the YOLOv4 detection network is proposed. First, the K-means++ algorithm was used to cluster the labelled data frames (ground truth) to determine the size of the anchor suitable for litchi. Then, the output size of the feature map of the original network was changed to make it more suitable for small-scale target detection. In addition, the images were preprocessed (cropped input) before they were fed into the network. To construct the litchi dataset, 400 images containing more than 20,000 targets were collected. Comparing the detection level to that of the original YOLOv4 model, the recall, precision, and F1 score values of the improved model increased from 0.81 to 0.825, 0.762 to 0.892, and 0.79 to 0.85, respectively. The experimental results indicate that the performance of the litchi detection method proposed in the study is significantly greater than the original model, and it meets the requirements for fruit monitoring in litchi orchards.