The Research Of Bars Automatically Counting Based On The Image Processing

Over the years, the numbering system of bars steel has been the problems which steel bars production enterprises do not resolve well. At present, companies mainly rely on manual measurement to complete the task of bar steel counting. Sometimes a group of bars on the need for regular counting and verification, this is heavy workload. And the long time counting work will cause worker physical exhaustion and more errors. Therefore, the development and research of bars automatically counting system is an urgent need to be resolved. It mainly study on the bar identification and count system from following parts in this paper, put up laboratory equipment similar to industrial environment, image preprocessing, comparative image by edge detection experiment, DALSA software redevelopment and the counting of objective image. Laboratory equipment in accordance with the industry to design, make it essential to achieve the standard. In the image preprocessing methods, it introduced processing methods based on mathematical morphology:removing image background noise、 image enhancement, image smoothing, image sharpening、image filtering. It makes the image preprocessing effect better. At the same time,it also resolved the uneven binary image caused by uneven illumination in the images collection process, which image enhancement, smoothing, sharpening and filtering are main problems in study. Provision to the image on the basis of equalization in the histogram, find the most suitable for transformation of the functions and increase contrast of the image in order to enhance the image.The main purpose of smoothing is remove noise.Sharpening is prominent to main objectives.And the main purpose of filtering eliminates shadows that are mainly caused by uneven light.This experiment is done in uniform lighting conditions.Presented in the bar section is similar gray value.And significantly difference with the background. Therefore, the method of image segmentation to binary image processing is used in the edge detection experiment. It uses a number of typical bar image edge detection methods to analyze. By experiment comparison, Canny algorithm is a more successful method.Whether noise removal or edge detection are very successful. Finally, the image edge detection to detect targets in the class circle.A new algorithm of Blob is used. It is a class of circle center to determine the class round the edge of the rectangular outline of the external characteristics such as the identification principle and recognition algorithms to meet certain requirements of independence of the regional round for the class object recognition. The paper also for the experimental and industrial-site problems often need to solve the problem are given and methods for the rapid identification and follow-up online bar laid the foundation for an accurate count.