The Application Research Of The Segmentation Of Near-Infrared Image Based On OTSU

Furnace is the critical equipment used in the oil refining, petrochemical, chemical fiber, chemical fertilizer industry. The tube plays the role of heating raw materials or the reactor. According to the Domestic and foreign using experience of furnace, unbalanced temperature distribution and frustration, over-heating problem in partial areas can easily occur due to the burner and gas supply, all of which pose great threat to the quality of products and production security. In order to deal with these problems and improve the quality of products, we need to conduct real-time monitoring over the temperature of furnace.Aiming at studying the application of the near-infrared of image segmentation based on OTSU, this article focuses on the imaging mechanism of infrared image, its image features, image processing technology (with emphasis on image segmentation), and theoretical knowledge related to the basic principles and OTSU algorithms program implementation. It basically grasps the principles and algorithm of image segmentation and carries out applied research and verify over the specific algorithm.The specific tasks of this article are as follows:1) To study the imaging mechanism of the Infrared image, and make a deep analysis and comparison of the various characteristics between the near-infrared images and visible images. To understand and master the near-infrared images compared with the shortcomings of ordinary visible light images, and analyze the important and difficult areas of near-infrared image segmentation, as well as proposing a two-dimensional OTSU proposed segmentation scheme.2) To carefully understand the definition and the principle of image segmentation, and study the types and a variety of algorithms theory of image segmentation and its evaluation methods. I carried out detailed comparative experiments over the advantages and disadvantages of several typical threshold segmentation algorithms.3) According to the characteristics of infrared image, it systematically studies the basic idea of the one-dimensional and the two-dimensional theory of OTSU algorithm, grasps the characteristics of the image segmentation by using the Matlab language to compile the two-dimensional OTSU algorithm. By comparing the two-dimensional with the one-dimensional OTSU the maximum entropy algorithm for image segmentation, it obtained the two-dimensional OTSU suitable for near-infrared image segmentation.