The Slab Tracking System Based On The Recognition Of The Characters

The slab is an important intermediate product in steel production process, logistics tracking for it helps to improve the level of enterprise information and automation. However, when the slab is dispatched from the converter shed to the hot-rolling mill, there is no effective method of tracking at present since it may be dealt with off-line processing in the convert shed.In this paper, a method based on character image recognition to track slab is proposed after analyzing the slab distribution process and existing information and automation equipment. Tracking slab delivery information automatically is realized by capturing image with the video camera fixed between the converter shed and hot-rolling mill and then recognizing slab characters based on digital image processing.Firstly, the difference image method is employed in order to judge whether the slab is in the field of camera vision; secondly, the binary image of slab flank is processed with vertical and horizontal projection to locate characters and segregate the single one; finally, Shape-Context is adopted to extract feature and Kuhn-Munkras algorithm is adopted to resolve the match problem of samples and templates, so slab characters are recognized. In image processing, a kind of adaptive binarization method is proposed according to the features of character image of slabs; and a kind of algorithm to eliminate interference spots based on connected domain mark is presented to remove the influence of oxidation spots. These measures reduce the interference and improve the quality of character image. In addition, voting mechanism and database validation strategy are developed to take full advantage of recognition results of multiple-frame image and the slab encoding information in the database, which improves the tracking performance.Experiment simulation proves the effectiveness of pretreatment, object detection, feature extraction and recognition algorithm. Further more, the slab information tracking system software is devised after installing the camera, industrial computer and the necessary transport equipment in a large domestic mills, which embeds tracking system into existing information and automation network. Actual results show that the recognition rate for slab characters of this system remains at a high level, which overcomes adverse factors, such as ever-changing of scene and light in the industrial site, instability of character quality and so on. The system realizes automatic tracking for slab information, improve the level of enterprise information and automation and guarantee the direct hot charging rolling.