An optimized YOLO‐based object detection model for crop harvesting system

Abstract

The adoption of automated crop harvesting system based on machine vision may improve productivity and optimize the operational cost. The scope of this study is to obtain visual information at the plantation which is crucial in developing an intelligent automated crop harvesting system. This paper aims to develop an automatic detection system with high accuracy performance, low computational cost and lightweight model. Considering the advantages of YOLOv3 tiny, an optimized YOLOv3 tiny network namely YOLO-P is proposed to detect and localize three objects at palm oil plantation which include fresh fruit bunch, grabber and palm tree under various environment conditions. The proposed YOLO-P model incorporated lightweight backbone based on densely connected neural network, multi-scale detection architecture and optimized anchor box size. The experimental results demonstrated that the proposed YOLO-P model achieved good mean average precision and F1 score of 98.68% and 0.97 respectively. Besides, the proposed model performed faster training process and generated lightweight model of 76 MB. The proposed model was also tested to identify fresh fruit bunch of various maturities with accuracy of 98.91%. The comprehensive experimental results show that the proposed YOLO-P model can effectively perform robust and accurate detection at the palm oil plantation. This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.

An automated solid waste detection using the optimized YOLO model for riverine management

“…Meanwhile, the one-stage detection method integrates region proposal and classification into one step, which reduces the detection time. The mainstream methods of one-stage detection are Single Shot Detector (SSD) and YOLO (27).…”

Fast and Accurate Object Detection in Remote Sensing Images Based on Lightweight Deep Neural Network

“…Compared with K-means clustering, the proposed algorithm can improve the detection accuracy by 1.13% in terms of mAP when using 9 anchors. (9,10), (15,15), (22,23), (31,34), (47,53), (145,178) K-means 9 82.00 (8,9), (12,12), (16,15), (19,20), (24,25), (30,32), (40,44), (51,58), (145,178) 6 81.69 (14,15), (24,27), (44,51), (96,180), (153,146), (187,213) Proposed 9 83.13 (13,13), (19,19), (30,31), (41,50), (50,60), (134,171), (99,197), (157,113), (176,214) Finally, to verify the generality of the proposed scheme, we have also tested the proposed detection framework on the DIOR dataset. As illustrated in Figure 10, the distribution of the obtained anchors by using the proposed scheme is also more evenly allocated than K-means clustering.…”

“…The IOU score is further used as the relative distance for anchor configurations in YOLOv4 [ 35 ]. Junos et al [ 47 ] used a similar scheme to optimize the anchor in YOLOv3 and applied the network to the crop harvesting system. Zlocha et al [ 48 ] optimized the anchor configuration based on a differential evolution search algorithm in RetinaNet.…”

Automated Data Annotation for 6-DoF AI-Based Navigation Algorithm Development

“…Object detection and instance segmentation algorithms are the pillars of many modern computer vision applications as they determine the location and totality of an object within an image, allowing one to carry out more complex tasks such as autonomous navigation [24,25], labor automation [26], and security surveillance [27,28].…”