基于改进 Deeplabv3+的电力线分割方法研究

针对已有的分割算法存在的复杂场景干扰大、分割不准确的问题，提出一种用于电力线分割任务的改进 Deeplabv3+ 模型。将原始主干网络替换为轻量级 Mobilenetv 2 网络，增加低水平特征，获得5路输入特征，充分提取特征信息；添加空洞 空间金字塔池化(atrous spatial pyramid pooling , ASPP )的卷积分支数量，调整空洞率，提升图像的特征抓取能力，进一步在每 个空洞卷积后加入1×1卷积操作，加快计算速度；提出一种基于坐标注意力机制的语义嵌入分支模块(coordinate attention semantic embedding branch , CASEB ),融合第2、3路特征，增强目标特征的表示；引入卷积注意力机制模块(convolution block attention module , CBAM )抑制无用信息的传递，提高模型识别效率。实验结果表明，相对于原 Deeplabv 3+模型，改进模型在 平均像素精度(mean pixel attention , MPA )和平均交并比(mean intersection over union , mloU )上分别提升2 . 37%和3 .42%, 该方法可提供更加精确的电力线分割结果。

Research on power line segmentation method based on improved Deeplabv3+

In order to solve the problems of complex scene interference , inaccurate segmentation and slow prediction ,an improved Deeplabv 3+ model for power line segmentation is proposed . Replace the original backbone network with lightweight Mobilenetv 2 network , add low -level features , obtain five -way input features , and fully extract feature information . The number of convolution in the atrous spatial pyramid pooling ( ASPP ) is increased , and the voidness rate was adjusted to improve the feature capturing ability of the image . Furthermore ,1×1 convolution operation was added after each void convolution to speed up the calculation . A coordinate attention semantic embedding branch ( CASEB) based on coordinate attention mechanism is proposed , which integrates the second and third features to enhance the representation of target features . CBAM attention module is introduced to inhibit the transmission of useless information and improve the efficiency of model recognition . Compared with the original Deeplabv 3+ model , the mean pixel attention ( MPA ) and mean intersection over union ( mloU ) of the improved model are improved by 2.37% and 3.42% respectively . This method can provide more accurate results of power line segmentation.