节点文献

基于改进Faster RCNN的变电站红外图像多目标识别

Multi-Target Recognition of Substation Infrared Image Based on Improved Faster RCNN

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 王妤陈秀新袁和金

【Author】 WANG Yu;CHEN Xiuxin;YUAN Hejin;Baoding Vocational & Technical College,Computer Information Engineering Department;North China Electric Power University,Computer Department;

【机构】 保定职业技术学院计算机信息工程系华北电力大学控制与计算机工程学院

【摘要】 为解决变电站红外图像中多种高压电气设备识别问题,文中提出一种基于改进Faster RCNN的变电站红外图像多目标识别方法。该方法通过VGG16提取红外图像中的多种电气设备图像特征,通过区域建议网络生成区域建议并通过边框回归调整区域建议,ROI Pooling将不同尺度的区域建议映射到尺寸固定的输出向量后送入Softmax进行分类,并按照区域建议包含关系对识别错误的部位类别进行修正。实验选取27 586张红外图像制作成VOC2007格式红外数据集,并对测试集中5 517张红外图像的识别结果进行统计,实验结果表明改进Faster RCNN识别准确率达到92.8%,较改进前提高了9.7%,具有较高的工程实用价值。

【Abstract】 In order to solve the problem of identifying multiple high-voltage electrical equipment in substation infrared images,this paper proposes a multi-target recognition method for substation infrared images based on improved Faster RCNN. This method uses VGG16 to extract image features of a variety of electrical equipment in infrared images. Generate regional suggestions through the regional suggestion network and adjust the regional suggestions through border regression. ROI Pooling maps regional suggestions of different scales to fixed-size output vectors and sends them to Softmax for classification,and corrects the incorrectly identified parts categories according to the inclusion relationship of the regional suggestions. In the experiment,27 586 infrared images are selected to make voc207 format infrared data set,and the recognition results of 5 517 infrared images in the test set are statistically analyzed. The experimental results show that the recognition accuracy of the improved fast RCNN is 92.8%,which is9.7% higher than that before,and has higher engineering practical value.

【基金】 中央高校基本科研业务费专项项目(2017MS157)
  • 【文献出处】 传感技术学报 ,Chinese Journal of Sensors and Actuators , 编辑部邮箱 ,2021年04期
  • 【分类号】TP391.41;TN219;TM63
  • 【被引频次】2
  • 【下载频次】420
节点文献中: 

本文链接的文献网络图示:

本文的引文网络