基于工业现场环境的图像增强方法研究
2019-04-04熊湾刘琼齐家栋
熊湾 刘琼 齐家栋



关键词: 工业环境; 图像增强; 混料颗粒; 直方图双向均衡化; 拉普拉斯变换; 高频提升滤波
中图分类号: TN911.73?34; TP391 文献标识码: A 文章编号: 1004?373X(2019)02?0163?05
Research on image enhancement method based on industrial site environment
XIONG Wan, LIU Qiong, QI Jiadong
(School of Information Science and Engineering, Wuhan University of Science and Technology, Wuhan 430081, China)
Abstract: In allusion to the weak and uneven illumination problems of the industrial environment, an image enhancement method combining the spatial domain and frequency domain is adopted to conduct enhancement processing of the shot mixed particle images. The basic principles of histogram bi?directional equalization, Laplacian operator and high?frequency emphasis filtering enhancement are discussed. During histogram bi?directional equalization, the grayscale processing is conducted from two directions of the gray density and gray interval of the histogram simultaneously, which can enhance the contrast of the whole image. The Laplace transform can sharpen images, and highlight the particle edge details so that the particles can be clearly presented. The high?frequency emphasis filtering is used to conduct smooth processing of the images after Fourier transform, so as to remove image noises. The experimental results obtained according to the above methods show that the enhancement effect of images enhanced using the new method is respectively 30%, 8 times and 84% better in aspects of image mean, mean square and contrast enhancement coefficient than the effect of original images, and has a distinct advantage over the effect obtained using a single certain enhancement method.
Keywords: industrial environment; image enhancement; mixed particle; histogram bi?directional equalization; Laplace transform; high?frequency emphasis filtering
0 引 言
在钢铁工业的继续发展和进步中,钢铁材料仍是最主要的结构和工程材料,其质量关系到建筑工程的结构质量和安全[1]。而在钢铁材料的冶炼中,天然富矿在产量和质量上都远远不能满足高炉冶炼的要求,大量贫矿经选矿后得到的精矿粉不能直接入炉冶炼,只能通过人工方法将这些粉矿制成块状的人造富矿供高炉使用。运用烧结法得到的烧结矿是我国高炉的主要原料,随着钢铁产量的日益增加,对烧结矿的质量要求越来越严苛[2]。烧结混合料的颗粒质量是影响烧结矿质量的主要因素之一,为了使烧结效果最佳,就需要在混料过程中对混料颗粒的分布进行严格控制,使烧结混料颗粒的质量保持在一定的规格内。为了更准确、更快速地分析混料质量,借助图像分析技术来检测混料。在实际工业现场采集图像时由于光照不均、照明度不高等各种原因导致拍摄的烧结混料颗粒图像存在亮度较低且颗粒模糊不清等。这样不仅会影响图像的视觉效果,而且会给后续图像的分割、特征提取等操作带来困难[3?4]。为了更好地识别分析图像中的颗粒,需对图像进行增强处理,其本质是通过一定的方法对原图像数据进行适当的变换,使增强后的图像更符合人的视觉效果[5?6],并提高颗粒和背景的对比度,使得颗粒更容易从背景中分离出来,并为进一步的图像分析打好基础。……
