Intelligent steel slag detection method and system based on convolutional neural network

Abstract

The invention discloses an intelligent steel slag detection method and system based on a convolutional neural network. The method comprises the steps of steel slag image recognition, steel flow targetdetection and color steel slag image segmentation, wherein a color steel slag image in a video frame image serves as an object, and the color steel slag image is recognized through an image recognition method based on an improved AlexNet convolutional neural network; the method comprises the following steps: detecting steel flow information in a color steel slag image, and detecting a steel flowfrom a complex background through a target detection method based on a YOLOv3 convolutional neural network, thereby accurately detecting the slag inclusion condition of the steel flow; a color image is preprocessed based on a K-means clustering algorithm of a Lab color space, and steel slag is completely separated from molten steel by adopting an improved Otsu image segmentation algorithm. And carrying out steel slag visual detection by utilizing a visual user interface system. The method is simple and easy to implement and low in cost, steel slag and molten steel can be distinguished, false detection is avoided, the real-time recognition precision of the steel slag image is improved, and the purity of the molten steel is improved.

Description

technical field [0001] The invention belongs to the technical field of steel slag detection, and in particular relates to an intelligent steel slag detection method and system based on a convolutional neural network. Background technique [0002] To produce high-quality, high-value-added steel, it is essential to strictly control steel slag entering the next process, such as hot metal desulfurization and slag removal and converter steelmaking are indispensable processes in the steelmaking process. Desulfurization and slag removal of molten iron is a pretreatment process for molten iron. The desulfurization process needs to add desulfurizer. After desulfurization, a layer of residue will float on the surface of the molten iron. Desulfurization of hot metal determines the level of sulfur content in molten iron at the end of treatment, and slag removal is an important means to remove high-sulfur slag after desulfurization treatment from molten iron, and is the main factor dete...

AI Technical Summary

Problems solved by technology

The human eye observation method has a large subjectivity and requires operators to have rich operating experience; the weighing detection method mainly uses the average value of the molten steel quality of multiple pours as a reference value to choose whether to close the nozzle, and the reference value will directly affect the detection. As a result, if the setting is too high, the slag content in the steel stream will exceed the standard, and if the setting is too low, it will lead to waste of resources; the electromagnetic detection method needs to install a special coil near the tapping hole. Due to the high temperature of the tapping hole, the service life of the coil is short, so the transformation and maintenance The cost is very high; the working environment temperature of the probe used by the ultrasonic detection method is as high as 1500°

Images