基于改进 U-Net 的高压电缆绝缘层图像分割研究

针对目前高压电缆绝缘层检测操作繁琐、效率低、重复测量差异大等问题,设计了一种新型电缆绝缘层检测装置,提出了一种基于改进 U-Net 的高压电缆绝缘层图像分割方法。 首先替换主干特征提取网络为 VGG16 网络,结合迁移学习将 VGG16在 Pascal VOC2012 数据集中训练的权重作为预训练权重,利用通道注意力模块在跳跃连接处融入自适应特征加权机制,在上采样过程中添加分组卷积,提高了语义分割精度;然后利用训练的最优权重进行绝缘层图像分割,提取轮廓区域特征并进行二值化处理,使用连通区域算法对轮廓区域进行填充;最后,融合原始图像和分割区域生成完整绝缘层分割图像。 实验结果表明,平均交并比和平均像素准确率达到 99. 56%和 99. 81%,较原网络效果提升明显,验证了该方法在高压电缆绝缘层分割上的有效性。

Research on image segmentation of high-voltage cables insulation layer based on improved U-Net

Aiming at the current problems of cumbersome operation, low efficiency and large variation in repeated measurements of high-voltage cable insulation layer quality inspection, a new type of cable insulation layer inspection device is designed, and a high-voltagecable insulation layer image segmentation method based on improved U-Net is proposed. Firstly, the backbone feature extraction networkis replaced with the VGG16 network, the weights trained by VGG16 in the Pascal VOC2012 dataset are used as the pre-training weightsin combination with the transfer learning, the adaptive feature weighting mechanism is incorporated in the jump connections by using thechannel attention module, as well as the grouped convolution is added in the up-sampling process, which improves the semanticsegmentation accuracy. Next, the insulating layer image segmentation is performed using the trained optimal weights, the contour regionfeatures are extracted and binarised, and the contour region is filled using the connected region algorithm. Finally, the completeinsulation layer segmentation image is generated by fusing the original image and the segmented region. The experimental results showthat the mean intersection-over-union and mean pixel accuracy reach 99. 56% and 99. 81%, which is a significant improvement over theoriginal network effect, and verifies the effectiveness of the method on the segmentation of the insulation layer of high-voltage cables.