High voltage cable is an important basic equipment for city power supply. Existing research shows that: the method of digital X-ray imaging can realize the non-destructive detection of power cable, it belongs to the new detection technology in the field of cable defect detection, at present, there is no deep X-ray digital image processing and defect recognition method, and it can not directly detect the cable ontology and defect recognition from the original digital image. Therefore, this article on the depth of power cable X-ray digital image processing and intelligent recognition technology carried out research, buffer layer defects using gray processing technology, the original image grayscale range compression to the human eye can recognize, and defect identification, and the traditional convolution neural network and the convolution neural network (CNN) FCN image data for training, implementation of the intelligent identification of power cable buffer layer defects. The results show that compared with traditional convolutional neural network (CNN), the full convolutional neural network (FCN) proposed in this paper has a clearer and more intuitive recognition effect.
[1]
Xiaogang Wang,et al.
Deep Learning Face Representation from Predicting 10,000 Classes
,
2014,
2014 IEEE Conference on Computer Vision and Pattern Recognition.
[2]
Vineeth N. Balasubramanian,et al.
ADINE: an adaptive momentum method for stochastic gradient descent
,
2017,
COMAD/CODS.
[3]
Ting Sun,et al.
Single-pixel imaging via compressive sampling
,
2008,
IEEE Signal Process. Mag..