Dragon fruit detection in natural orchard environment by integrating lightweight network and attention mechanism

Abstract

An improved lightweight network (Improved YOLOv5s) was proposed based on YOLOv5s in this study to realise all-weather detection of dragon fruit in a complex orchard environment. A ghost module was introduced in the original YOLOv5s to realise the lightweight of the model. The coordinate attention mechanism was joined to make the model accurately locate and identify the dense dragon fruits. A bidirectional feature pyramid network was built to improve the detection effect of dragon fruit at different scales. SIoU loss function was adopted to improve the convergence speed during model training. The improved YOLOv5s model was used to detect a dragon fruit dataset collected in the natural environment. Results showed that the mean average precision (mAP), precision (P) and recall (R) of the model was 97.4%, 96.4% and 95.2%, respectively. The model size, parameters (Params) and floating-point operations (FLOPs) were 11.5 MB, 5.2 M and 11.4 G, respectively. Compared with the original YOLOv5s network, the model size, Params and FLOPs of the improved model was reduced by 20.6%, 18.75% and 27.8%, respectively. Meanwhile, the mAP of the improved model was improved by 1.1%. The results prove that the improved model had a more lightweight structure and better detection performance. Moreover, the average precision (AP) of the improved YOLOv5s for dragon fruit under the front light, back light, side light, cloudy day and night was 99.5%, 97.3%, 98.5%, 95.5% and 96.1%, respectively. The detection performance met the requirements of all-weather detection of dragon fruit and the improved model had good robustness. This study provides a theoretical basis and technical support for fruit monitoring based on unmanned aerial vehicle technology and intelligent picking based on picking robot technology.

Improving insulator fault detection with effective-YOLOv7 network

“…Model a used the YOLOv7 model in its original form, whereas model b replaced the original loss function with SIoU (SCYLLA-IoU) to demonstrate the effectiveness of the bounding box loss function replacement proposed in this study. The SIoU regression loss function 32 redefines the penalty measure by considering the vector angles between required regressions, which can greatly accelerate the training convergence process. Literature 33 has shown through experiments that SIoU can speed up network convergence.…”

A Dragon Fruit Picking Detection Method Based on YOLOv7 and PSP-Ellipse

“…However, this method makes the average precision of the model decrease slightly [21]. Zhang et al [22] proposed an improved YOLOv5s to accurately detect dragon fruits under different light conditions. These methods above just locate and classify the fruits, which is difficult to apply to picking just some fruits, such as dragon fruit.…”

A simplified network topology for fruit detection, counting and mobile-phone deployment

“…The YOLO-Jujube introduced by Xu et al [ 21 ] to detect jujube fruit automatically for ripeness inspection achieved AP of 88.8% and a speed of 245 fps, and Yan et al [ 22 ] recorded AP of 86.75% and a speed of 66.7 fps to detect apple fruit targets based on improved YOLOv5. To detect a dragon fruit in the natural environment, Zhang et al [ 23 ] realized AP of 97.4% with the incorporation of ghost network (Han et al [ 24 ]) into YOLOv5. Qiao et al [ 6 ] introduced ShuffleNetv2 network proposed by Ma et al [ 25 ] into YOLOv5 for a counting method of red jujube, which noted AP of 94% and speed of 35.5 fps.…”

“…The recent YOLOv7 designed by Wang et al [ 30 ] was declared to have surpassed YOLOv4 and YOLOv5 in detection performance. For this reason, Zhang et al [ 23 ] applied both YOLOv7 and YOLOv7-tiny for dragon fruit detection, and respectively realized AP of 95.6% and 96.0%. Additionally, Chen et al [ 31 ] improved the YOLOv7 using a CBAM (Convolutional Block Attention Module) for citrus detection, achieving AP of 97.29% and a speed of 14.4 fps.…”