一种改进的压板状态识别SSD算法

An improved SSD algorithm based on pressure plate status recognition

在变电站二次侧管理中,压板承担着重要作用。文章提出了一种改进SSD图像识别算法,用以实现对压板状态的识别。新算法通过在SSD目标识别算法中,嵌入注意力机制,利用注意力机制挖掘了每个特征通道的重要程度,提升有用特征的权重,抑制了无效特征,提升了原有算法的检测精度。为了解决训练样本不足的问题,新算法通过对样本的扩充和迁移学习的方式,训练得到了提出的新SSD算法中的各个参数,并通过仿真实验进行验证。实验结果表明,改进后的SSD算法,其识别准确率达到96%,召回率达到94%,每秒可以检测23张图片,能够有效提升变电站内压板状态识别的效率。

In the secondary side management of the substation,the pressure plate plays a major role.This paper proposes an improved SSD image recognition algorithm to realize the recognition of the pressure state of the plate.The novel algorithm embeds the attention mechanism in the SSD target recognition algorithm,utilizes the attention mechanism to mine the importance of each feature channel,increases the weight of useful features,suppresses invalid features,and improves the detection accuracy of the original algorithm.In order to solve the problem of insufficient training samples,the novel algorithm trains the parameters of the proposed novel SSD algorithm by means of sample expansion and migration learning,which is verified by simulation experiments.Experimental results show that the improved SSD algorithm has a recognition accuracy rate of 96%,a recall rate of 94%,and 23 images per second can be detected,which can effectively improve the efficiency of the pressure plate status in the substation.