Edge Detection Method For Jar Image Recognition

With the rapid economic development,the demand for petroleum has increased substantially.As a major transportation tool for petroleum transportation,the railroad tanker has become an important operation process for oil companies.This process is very tedious.The original manual positioning and filling device has a very high requirement for the operator,and the labor intensity is high.This can easily cause fatigue operations and safety accidents.The positioning of the tank mouth is difficult to be precise,and together with the low working efficiency it is difficult to meet the needs of the enterprise.Therefore,we need urgently to improve the existing methods for the positioning of irrigation ports.In addition,the existing tank positioning system for railway tank trucks which have been put into use already adopts machine vision positioning technology generally and PLC control technology to collect and process images of tanker tank ports and perform automatic positioning.This kind of positioning system has a relatively high requirement for the stability of the environment,such as the change of light,the increase of dust in the air,the oil on the body of the tanker,etc.These factors will affect its recognition rate.In order to solve the above problems,this paper would use machine vision technology and image processing technology to study a new edge detection method for image recognition of tank cars comprehensively.The main work includes the following aspects:First,we designed and implemented an automatic tank positioning system for oil tankers.The system included a tanker drag subsystem,a big crane tube movement subsystem,a irrigation mouth identification subsystem,and a control subsystem.The image acquisition of the irrigation port identification subsystem adopted industrial detection cameras and FA lenses,and image processing used industrial computers to save costs and improve system flexibility.Visual image system and PLC interface would be used for motion control.Secondly,an edge detection method would be designed for image recognition of tank cars.The method was based on the Canny edge detection operator,which extracted the image of the edge of the tanker can,preprocessed the image,extracted the feature via the edge detection and the optimal threshold algorithm,and then adopted the image template matching method to identify the cans and determine the position of tanker irrigation port quickly.Finally,this article would be based on the realization of the auto-positioning system of tanker irrigation port,and some related experiments would be designed to test the method of edge detection for image recognition of tanker irrigation port.Experimental results could verify the feasibility and correctness of the proposed method.