激光与光电子学进展, 2021, 58 (2): 0215009, 网络出版: 2021-01-11
基于改进U-Net卷积神经网络的钢轨表面损伤检测方法 下载: 1132次
Rail Surface Damage Detection Method Based on Improved U-Net Convolutional Neural Network
图 & 表
图 2. 不同操作扩展后的样本图像。(a)原始图像;(b)注释图像;(c)平移变换;(d)旋转变换;(e)缩放变换
Fig. 2. Sample images expanded by different operations. (a) Original images; (b) ground truth; (c) translation transformation; (d) rotation transformation; (e) scaling transformation
图 4. 所提方法在模型训练过程中的性能曲线
Fig. 4. Performance curves of proposed method in model training process
梁波, 卢军, 曹阳. 基于改进U-Net卷积神经网络的钢轨表面损伤检测方法[J]. 激光与光电子学进展, 2021, 58(2): 0215009. Bo Liang, Jun Lu, Yang Cao. Rail Surface Damage Detection Method Based on Improved U-Net Convolutional Neural Network[J]. Laser & Optoelectronics Progress, 2021, 58(2): 0215009.