Research On Grayscale Method In Container Number Recognition And Location

In recent years,the rapid economic growth has continuously promoted the development of the logistics industry,and container transportation has become one of the main ways of logistics transportation.In order to effectively manage containers,the intelligent container identification system based on machine vision has been widely used.Before the intelligent container identification system recognizes the collected container image,the image needs to be grayed out,and the image data after the graying is small and contains most of the information of the original image,which does not affect the subsequent identification of the box number.Moreover,the operation speed can be reduced.However,in the process of graying,due to the noise introduced in the transmission process or the contamination of the container character area,the image contrast is low or missing information,resulting in low recognition rate and recognition slow speed.In this paper,the intelligent container identification system is used to identify the container number,and the related research work is carried out around the image graying method.Aiming at the noise introduced in the transmission of the intelligent container identification system and the shadow in the image character area,a gray image optimization method based on gray stretching combined with gray histogram equalization is proposed.This method uses K-means algorithm can enhance the region contrast by analyzing the gray level distribution of the image and changing the gray level of the region,so as to reduce the noise produced in the gray scale process.Meanwhile,the details of the grayscale image of the color image are retained.Aiming at the containers character recognition area contamination and other problems,a hybrid method of principal component analysis(PCA)combined with Bayesian threshold estimation of gray change rate is proposed to optimize the gray level of the image.It can make up for the missing information and effectively determine the edge features after judging the change rate of the gray value of a certain point in the image and the gray value of the neighboring pixel points,thus greatly improving the character recognition accuracy of the subsequent sequence.Finally,a container intelligent identification system for logistics parks is realized by evaluating the comparison and selecting the principal component analysis method combined with the Bayesian threshold to estimate the gray rate of change.Experiments show that in the recognition of 78 container numbers in logistics park,the accuracy of single character and box number can reach 95% and 98% respectively by using the hybrid gray method proposed in this paper,compared with the common mean method and weighted average method.