Research of status recognition of Fiber transfer box based on machine vision and deep learning
- PDF / 1,795,830 Bytes
- 15 Pages / 439.37 x 666.142 pts Page_size
- 54 Downloads / 160 Views
Research of status recognition of Fiber transfer box based on machine vision and deep learning Meng Li 1
1
1
& Yinhui Ao & Wenzheng Peng & Jinghui He
1
Received: 4 September 2019 / Revised: 21 June 2020 / Accepted: 9 July 2020 # Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract
Fiber transfer box is a box containing fiber ports which transfer optical signals for telecommunication networks. It’s hard to check and record the status of hundreds of ports using man power. This paper proposes an intelligent recognition and localization system for fiber transfer box. Pictures of ports are treated and classified with Support Vector Machine (SVM) and Deep learning algorithm. It can identify the status of all ports in the fiber transfer box and recognize the writing letter for each layer. First, the image is converted to binary one by threshold of the Cr channel. A SVM classifier is used to identify the Red-Hat ports by HOG features. Second, an adaptive chessboard segmentation algorithm is designed for segmentation of all ports and character area. In order to improve the identification accuracy, an eleven-layer Convolution Neural Network (CNN) is trained and used to further identify the ports which have not been classified correctly by SVM. Letters are extracted and positioned by using YOLOv3-tiny network. Experiments show that the method achieves great accuracy and efficiency for a variety of scenes on mobile devices. Keywords Support Vector Machine (SVM) . You Look Only Once (YOLO) . Adaptive Checkerboard Segmentation . Convolution Neural Network (CNN)
1 Introduction With the wide application of fiber network, fiber transfer boxes spread around streets and buildings. In order to keep the transfer box function normal working, Telecommunication Company will send workers to check and correct the status of fiber box. Traditional maintaining is to check and record the status of each port. The disordered and unsystematic fiber lines and port shapes may cause a huge workload for maintenance jobs, errors and delays are often happened in
* Meng Li [email protected]
1
School of Mechanical and Electrical Engineering, Guangdong University of Technology, Guangzhou 510006, China
Multimedia Tools and Applications
this occasion. So, a new checking way is to take pictures of fiber box and the images of fiber ports are treated to give out the status of each port automatically. It’s more convenient and efficient with the machine vision technology to identify fiber port status in these maintenance jobs. Identification and classification involve two stages: feature extraction and model classification [9, 21]. Mitraet al. [12] presented a neural network architecture using a Support Vector Machine as an inference engine for classification of light detection. Zhang et al. [22] selected features of Windows Registry access recorder, and implemented the standard SVM, weighted SVM and one class SVM for status detection. In the classification process, the speed and accuracy were often degraded by undistinguished
Data Loading...
