Port Container Number Recognition System Based on Improved YOLO and CRNN Algorithm
Authors:
- 1 June 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 56, 72-77
Abstract
Port container number recognition algorithm, as an application example of OCR technology, is developing rapidly. However, the traditional container number identification algorithm from image preprocessing to location and then segmentation is not only becoming more and more time-consuming, but also has many disadvantages. At present, most of the deep learning algorithms are multi-stage, which need multi-stage adjustment in training. In order to solve the above problems, this paper proposes a container number recognition algorithm based on improved lightweight YOLOV3 and improved CRNN. That is, on the basis of the original algorithm, by reducing the number of parameters and the number of residuals in each layer of the original network, we use the deep separable convolution, introduce the Inverted Residuals block of MobileNetV2, and reorganize residual blocks to get lightweight. The end to end recognition algorithm has the advantages of high accuracy and fast speed, the accuracy of case recognition is 96%.Keywords
This publication has 10 references indexed in Scilit:
- Light-Weight Object Detection Networks for Embedded PlatformActa Optica Sinica, 2019
- Towards End-to-End Text Spotting with Convolutional Recurrent Neural Networks2017
- Comparison of SVM and Naïve Bayes Text Classification Algorithms using WEKAInternational Journal of Engineering Research & Technology, 2017
- Densely Connected Convolutional Networks2017
- YOLO9000: Better, Faster, Stronger2017
- Faster R-CNN: Towards Real-Time Object Detection with Region Proposal NetworksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2016
- Deep Residual Learning for Image Recognition2016
- You Only Look Once: Unified, Real-Time Object Detection2016
- Convolutional Neural Networks for Sentence Classification2014
- Long Short-Term MemoryNeural Computation, 1997