FeO content of sinter is an important indicator of the quality of sinter. Aiming to overcome the difficulty of detecting the FeO content of sinter in the sintering process in real-time, this paper proposes a polymorphic measurement method for sinter FeO content based on heterogeneous features of infrared thermal images. First, an infrared thermal imager is applied to capture the infrared thermal images of sinter cross section at the tail of the sintering machine, and key frame and region of interest extraction are adopted to reduce the data throughput. Then, the shallow features and deep features that are related to the FeO content are extracted based on the regions of interest. Next, a polymorphic mechanism model is established to obtain the preliminary FeO content, and the sinter quality is divided into three grades according to the preliminary FeO content. Finally, three intelligent models corresponding to the three sinter grades are established to achieve the FeO content prediction based on the extracted heterogeneous features. Results in a sintering plant show that the proposed method can measure the FeO content accurately and provide reliable FeO content data for sintering site.

中文翻译：

---

基于红外热图像异质性的烧结矿FeO含量多态性测量方法

烧结矿中的FeO含量是衡量烧结矿质量的重要指标。为了克服烧结过程中实时检测烧结矿中FeO含量的困难，提出了一种基于红外热图像异质性特征的烧结矿FeO含量多态性测量方法。首先，采用红外热像仪在烧结机尾部捕获烧结截面的红外热像，并采用关键帧和感兴趣区域提取以降低数据吞吐量。然后，基于感兴趣区域提取与FeO含量有关的浅层特征和深层特征。接下来，建立多态机制模型以获得初步的FeO含量，根据初步的FeO含量将烧结矿质量分为三个等级。最后，建立了与三个烧结矿等级相对应的三个智能模型，以基于提取的异质性特征实现FeO含量的预测。烧结厂的结果表明，所提出的方法可以准确地测量FeO的含量，并为烧结现场提供可靠的FeO含量数据。