基于E-FCNN的电力巡检图像增强

为了解决无人机巡线、无人值守变电站机器人巡检中，由于距离过远或机器抖动造成的采集图像待检目标分辨率低、图像模糊等问题，提出一种边缘感知反馈卷积神经网络E-FCNN。该网络在传统超分辨率网络基础上增加了残差模块和反馈机制，实现细节特征的提取和强化，并通过边缘感知分支补充纹理信息，提升了图像的细节描述。通过测试集实验结果表明：提出的边缘感知反馈卷积神经网络无论在主观视觉质量，或是峰值信噪比等客观评价指标上，都明显优于其他相关算法。且在基于无人机巡检的绝缘子检测应用中能够有效提高绝缘子检测率。

E-FCNN Based Electric Power Inspection Image Enhancement

For UAV patrol of transmission lines and robot inspection of unattended substations, low image resolution is one of the main problems due to long shooting distance or machine shaking. In order to solve this problem, we propose an edge-aware feedback convolutional neural network (E-FCNN), which not only adds Resnet blocks and feedback mechanism to the conventional super-resolution network to strengthen the ability of feature extraction, but also adds texture information to the edge-aware branch to enhance the image detail. Extensive experiments show that the proposed algorithm is superior to other existing algorithms, both in subjective visual quality and objective evaluation indexes such as peak signal-to-noise ratio. Practically, the proposed algorithm can improve the accuracy of insulator detection in UAV transmission line inspection.