Introduction: Fruit detection is one of the key functions of an automatic picking robot, but fruit detection accuracy is seriously decreased when fruits are against a disordered background and in the shade of other objects, as is commmon in a complex orchard environment.Methods: Here, an effective mode based on YOLOv5, namely YOLO-P, was proposed to detect pears quickly and accurately. Shuffle block was used to replace the Conv, Batch Norm, SiLU (CBS) structure of the second and third stages in the YOLOv5 backbone, while the inverted shuffle block was designed to replace the fourth stage’s CBS structure. The new backbone could extract features of pears from a long distance more efficiently. A convolutional block attention module (CBAM) was inserted into the reconstructed backbone to improve the robot’s ability to capture pears’ key features. Hard-Swish was used to replace the activation functions in other CBS structures in the whole YOLOv5 network. A weighted confidence loss function was designed to enhance the detection effect of small targets.Result: At last, model comparison experiments, ablation experiments, and daytime and nighttime pear detection experiments were carried out. In the model comparison experiments, the detection effect of YOLO-P was better than other lightweight networks. The results showed that the module’s average precision (AP) was 97.6%, which was 1.8% higher than the precision of the original YOLOv5s. The model volume had been compressed by 39.4%, from 13.7MB to only 8.3MB. Ablation experiments verified the effectiveness of the proposed method. In the daytime and nighttime pear detection experiments, an embedded industrial computer was used to test the performance of YOLO-P against backgrounds of different complexities and when fruits are in different degrees of shade.Discussion: The results showed that YOLO-P achieved the highest F1 score (96.1%) and frames per second (FPS) (32 FPS). It was sufficient for the picking robot to quickly and accurately detect pears in orchards. The proposed method can quickly and accurately detect pears in unstructured environments. YOLO-P provides support for automated pear picking and can be a reference for other types of fruit detection in similar environments.
It is necessary to detect multi-type farmland obstacles in real time and accurately for unmanned agricultural vehicles. An improved YOLOv5s algorithm based on the K-Means clustering algorithm and CIoU Loss function was proposed to improve detection precision and speed up real-time detection. The K-Means clustering algorithm was used in order to generate anchor box scales to accelerate the convergence speed of model training. The CIoU Loss function, combining the three geometric measures of overlap area, center distance and aspect ratio, was adopted to reduce the occurrence of missed and false detection and improve detection precision. The experimental results showed that the inference time of a single image was reduced by 75% with the improved YOLOv5s algorithm; compared with that of the Faster R-CNN algorithm, real-time performance was effectively improved. Furthermore, the mAP value of the improved algorithm was increased by 5.80% compared with that of the original YOLOv5s, which indicates that using the CIoU Loss function had an obvious effect on reducing the missed detection and false detection of the original YOLOv5s. Moreover, the detection of small target obstacles of the improved algorithm was better than that of the Faster R-CNN.
It is crucial for robotic picking fruit to recognize fruit accurately in orchards, this paper reviews the applications and research results of target recognition in orchard fruit picking by using machine vision and emphasizes two methods of fruit recognition: the traditional digital image processing method and the target recognition method based on deep learning. Here, we outline the research achievements and progress of traditional digital image processing methods by the researchers aiming at different disturbance factors in orchards and summarize the shortcomings of traditional digital image processing methods. Then, we focus on the relevant contents of fruit target recognition methods based on deep learning, including the target recognition process, the preparation and classification of the dataset, and the research results of target recognition algorithms in classification, detection, segmentation, and compression acceleration of target recognition network models. Additionally, we summarize the shortcomings of current orchard fruit target recognition tasks from the perspectives of datasets, model applicability, universality of application scenarios, difficulty of recognition tasks, and stability of various algorithms, and look forward to the future development of orchard fruit target recognition.
In order to remedy the defects of single sensor in robustness, accuracy, and redundancy of target detection, this paper proposed a method for detecting obstacles in farmland based on the information fusion of a millimeter wave (mmWave) radar and a camera. Combining the advantages of the mmWave radar in range and speed measurement and the camera in type identification and lateral localization, a decision-level fusion algorithm was designed for the mmWave radar and camera information, and the global nearest neighbor method was used for data association. Then, the effective target sequences of the mmWave radar and the camera with successful data association were weighted to output, and the output included more accurate target orientation, longitudinal speed, and category. For the unassociated sequences, they were tracked as new targets by using the extended Kalman filter algorithm and were processed and output during the effective life cycle. Lastly, an experimental platform based on a tractor was built to verify the effectiveness of the proposed association detection method. The obstacle detection test was conducted under the ROS environment after solving the external parameters of the mmWave radar and the internal and external parameters of the camera. The test results show that the correct detection rate of obstacles reaches 86.18%, which is higher than that of a single camera with 62.47%. Furthermore, through the contrast experiment of the sensor fusion algorithms, the detection accuracy of the decision level fusion algorithm was 95.19%, which was higher than 4.38% and 6.63% compared with feature level and data level fusion, respectively.