Rapid detection of Yunnan Xiaomila based on lightweight YOLOv7 algorithm

Abstract

IntroductionReal-time fruit detection is a prerequisite for using the Xiaomila pepper harvesting robot in the harvesting process.MethodsTo reduce the computational cost of the model and improve its accuracy in detecting dense distributions and occluded Xiaomila objects, this paper adopts YOLOv7-tiny as the transfer learning model for the field detection of Xiaomila, collects images of immature and mature Xiaomila fruits under different lighting conditions, and proposes an effective model called YOLOv7-PD. Firstly, the main feature extraction network is fused with deformable convolution by replacing the traditional convolution module in the YOLOv7-tiny main network and the ELAN module with deformable convolution, which reduces network parameters while improving the detection accuracy of multi-scale Xiaomila targets. Secondly, the SE (Squeeze-and-Excitation) attention mechanism is introduced into the reconstructed main feature extraction network to improve its ability to extract key features of Xiaomila in complex environments, realizing multi-scale Xiaomila fruit detection. The effectiveness of the proposed method is verified through ablation experiments under different lighting conditions and model comparison experiments.ResultsThe experimental results indicate that YOLOv7-PD achieves higher detection performance than other single-stage detection models. Through these improvements, YOLOv7-PD achieves a mAP (mean Average Precision) of 90.3%, which is 2.2%, 3.6%, and 5.5% higher than that of the original YOLOv7-tiny, YOLOv5s, and Mobilenetv3 models, respectively, the model size is reduced from 12.7 MB to 12.1 MB, and the model’s unit time computation is reduced from 13.1 GFlops to 10.3 GFlops.DiscussionThe results shows that compared to existing models, this model is more effective in detecting Xiaomila fruits in images, and the computational complexity of the model is smaller.

Optimizing the YOLOv7-Tiny Model with Multiple Strategies for Citrus Fruit Yield Estimation in Complex Scenarios

“…Ma et al [18] introduced a BiFPN module based on YOLOv7-tiny for detecting apples under different weather conditions, with an accuracy of 80.1%. Wang et al [19] replaced the regular convolution in the YOLOv7-tiny backbone network by using variability convolution and introducing an SE attention mechanism for detecting millet chili at different maturity levels, with a model accuracy of 90.3%. In the above studies, most of the networks focused on local area feature extraction and matching, and had a lightweight character while losing the semantic information of the target fruits.…”

Yolo-Chili: An Efficient Lightweight Network Model for Localization of Pepper Picking in Complex Environments

“…The success of the above methods proves the success of target detection in the field of fruit picking, but due to the problems of dense growth of chili fruits, uneven fruit size, severe occlusion of fruits by branches and leaves and similar backgrounds in chili pepper picking, it is difficult to target chili pepper fruits for efficient picking with the above methods [25][26][27][28][29][30][31][32]. At the same time, some of the current general-purpose models have problems such as insufficient model detection performance, large environmental interference factors, large model structure, and slow inference speed.…”