Container number identification method and device

Abstract

The invention discloses a container number identification method and device. The method comprises the following steps: 1) inputting an original picture of a container; 2) carrying out container surface coarse positioning by using an image semantic segmentation model, segmenting a container in the image, and carrying out image preprocessing to obtain a container type image; 3) conducting text detection on the container type image through a CPTN model; 4) conducting OCR recognition on all results of text detection through a CRNN model, and 5) regularizing recognition results according to a container number format and outputting a final container number.

Description

technical field [0001] The invention relates to the field of containers, in particular to a container number identification method and device. Background technique [0002] With the advancement of science and technology and the huge growth of trade among countries around the world, the logistics and transportation industry has accelerated its development year by year, and the global shipping volume has hit new highs year by year. As the main transportation and loading method of ocean transportation, containers play a very important role in ocean transportation. While shipping is booming, the management of containers requires more modern means to achieve the purposes of tracking a huge number of containers, cargo transfer and cargo ship storage planning. [0003] The container number is the unique identification code of each container. Automatic identification of the container number is the basis for various efficient management operations on the container, such as: affiliat...

AI Technical Summary

Problems solved by technology

However, since there is no unified standard for the automatic identification of container numbers in the world, image processing technology is mainly used to realize the automatic identification of container numbers at present.

However, the wrongly identified containers can only be recorded and checked by manual identification, which undoubtedly increases labor costs and transit time

[0004] At present, the box number recognition system cannot completely segment the incomplete and glued characters in the box number; at the same time, the recognition accuracy of the segmented incomplete characters is low

However, for incorrectly identified containers, it is still only possible to record and verify the container number through manual identification, which undoubtedly increases labor costs and circulation time

The existing box number recognition technology uses a single character for recognition, and the recognition speed is slow; and some images with box numbers have a certain shooting angle, and the recognition accuracy is low

Images