The accurate detection of green citrus in natural environments is a key step in realizing the intelligent harvesting of citrus through robotics. At present, the visual detection algorithms for green citrus in natural environments still have poor accuracy and robustness due to the color similarity between fruits and backgrounds. This study proposed a multi-scale convolutional neural network (CNN) named YOLO BP to detect green citrus in natural environments. Firstly, the backbone network, CSPDarknet53, was trimmed to extract high-quality features and improve the real-time performance of the network. Then, by removing the redundant nodes of the Path Aggregation Network (PANet) and adding additional connections, a bi-directional feature pyramid network (Bi-PANet) was proposed to efficiently fuse the multilayer features. Finally, three groups of green citrus detection experiments were designed to evaluate the network performance. The results showed that the accuracy, recall, mean average precision (mAP), and detection speed of YOLO BP were 86, 91, and 91.55% and 18 frames per second (FPS), respectively, which were 2, 7, and 4.3% and 1 FPS higher than those of YOLO v4. The proposed detection algorithm had strong robustness and high accuracy in the complex orchard environment, which provides technical support for green fruit detection in natural environments.
The accurate detection and counting of fruits in natural environments are key steps for the early yield estimation of orchards and the realization of smart orchard production management. However, existing citrus counting algorithms have two primary limitations: the performance of counting algorithms needs to be improved, and their system operation efficiency is low in practical applications. Therefore, in this paper, we propose a novel end-to-end orchard fruit counting pipeline that can be used by multiple unmanned aerial vehicles (UAVs) in parallel to help overcome the above problems. First, to obtain on-board camera images online, an innovative UAV live broadcast platform was developed for the orchard scene. Second, for this challenging specific scene, a detection network named Citrus-YOLO was designed to detect fruits in the video stream in real-time. Then, the DeepSort algorithm was used to assign a specific ID to each citrus fruit in the online UAV scene and track the fruits across video frames. Finally, a nonuniform distributed counter was proposed to correct the fruit count during the tracking process, and this can significantly reduce the counting errors caused by tracking failure. This is the first work to realize online and end-to-end counting in a field orchard environment. The experimental resultsshow that the F1 score and mean absolute percentage error of the method are 89.07% and 12.75%, respectively, indicating that the system can quickly and accurately achieve fruit counting in large-scale unstructured citrus orchards.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.