Abstract:

YOLO v3 object detection algorithm has been widely used in industry due to its high speed and high precision. However, there is a problem that the object function is not consistent with the evaluation index. To solve this problem, improved YOLO v3 object detection algorithm is proposed. This algorithm improves the GIoU calculation method and combines it with the YOLO v3 algorithm’s objective function to design a new objective function to achieve IoU local optimization as the local optimization of the objective function. The test results of public dataset VOC2007 and helmet wearing dataset show that the mAP-50 of the improved YOLO v3 increased by 2.07% and 2.05%, respectively, compared with the YOLO v3 algorithm.

Key words: object detection, YOLO v3 algorithm, GIoU algorithm, helmet wearing test

摘要：

YOLO v3目标检测算法由于其速度快、精度较高，在工业中获得了广泛应用，但存在目标函数与评价指标不统一的问题，针对此问题提出了改进YOLO v3目标检测算法。该算法改进GIoU计算方法，并与YOLO v3算法目标函数相结合，设计了一个新的目标函数，实现了目标函数局部最优为IoU局部最优。公共数据集VOC2007和安全帽佩戴数据集测试结果表明，相比于YOLO v3算法，改进YOLO v3的mAP-50分别提高了2.07%和2.05%。

关键词: 目标检测, YOLO v3算法, GIoU算法, 安全帽佩戴检测