基于模拟退火遗传算法的炉膛火焰图像分割

Image Segmentation for Furnace Flame based on SAGA

煤粉锅炉的燃烧过程是一个复杂而不稳定的悬浮燃烧过程,所以炉膛火焰图像的分割是存在一定的困难。本文介绍了遗传算法的基本原理,并结合OTSU最优阈值图像分割方法,提出了一种利用遗传算法进行图像分割的算法。同时,为了提高算法的运行性能,本文使用模拟退火方法对原算法进行了改进,并得到了更好的应用效果。

Process of burning coal powder in the boiler is complex suspension combustion and very unstable, so the image segmentation of the flame is quit difficult. Firstly in this paper, the principles of Genetic Algorithms are introduced. Combined with the OTSU and GA, a new image segmentation method is discussed. In addition, SAGA has been used to improve the appearance of the algorithm, and it makes a better result.